Red Hat Directory Server Administration Guide

Red Hat Directory Server 8.2
Administration Guide
for managing Directory Server instances
Edition 8.2.8
Landmann
Red Hat Directory Server 8.2 Administration Guide
for managing Directory Server instances
Edition 8.2.8
Landmann
[email protected] m
Legal Notice
Copyright © 2010 Red Hat, Inc..
T his document is licensed by Red Hat under the Creative Commons Attribution-ShareAlike 3.0 Unported
License. If you distribute this document, or a modified version of it, you must provide attribution to Red
Hat, Inc. and provide a link to the original. If the document is modified, all Red Hat trademarks must be
removed.
Red Hat, as the licensor of this document, waives the right to enforce, and agrees not to assert, Section
4d of CC-BY-SA to the fullest extent permitted by applicable law.
Red Hat, Red Hat Enterprise Linux, the Shadowman logo, JBoss, MetaMatrix, Fedora, the Infinity Logo,
and RHCE are trademarks of Red Hat, Inc., registered in the United States and other countries.
Linux ® is the registered trademark of Linus T orvalds in the United States and other countries.
Java ® is a registered trademark of Oracle and/or its affiliates.
XFS ® is a trademark of Silicon Graphics International Corp. or its subsidiaries in the United States
and/or other countries.
MySQL ® is a registered trademark of MySQL AB in the United States, the European Union and other
countries.
Node.js ® is an official trademark of Joyent. Red Hat Software Collections is not formally related to or
endorsed by the official Joyent Node.js open source or commercial project.
T he OpenStack ® Word Mark and OpenStack Logo are either registered trademarks/service marks or
trademarks/service marks of the OpenStack Foundation, in the United States and other countries and
are used with the OpenStack Foundation's permission. We are not affiliated with, endorsed or
sponsored by the OpenStack Foundation, or the OpenStack community.
All other trademarks are the property of their respective owners.
Abstract
T his book is for LDAP administrators.
Table of Contents
Table of Contents
.Preface
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
............
1. Directory Server Overview
18
2. Examples and Formatting
18
2.1. Command and File Examples
18
2.2. T ool Locations
19
2.3. LDAP Locations
19
2.4. T ext Formatting and Styles
19
3. Additional Reading
20
4. Giving Feedback
21
5. Documentation History
21
.Chapter
. . . . . . . . 1.
. . .Basic
. . . . . . Red
. . . . .Hat
. . . .Directory
. . . . . . . . . .Server
. . . . . . . Settings
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
............
1.1. System Requirements
24
1.1.1. Required JDK
24
1.1.2. Directory Server Supported Platforms
24
1.1.3. Directory Server Console Supported Platforms
25
1.1.4. Password Sync Service Platforms
25
1.1.5. Web Application Browser Support
25
1.2. Directory Server File Locations
25
1.3. LDAP T ool Locations
26
1.4. Starting and Stopping Servers
27
1.4.1. Starting and Stopping Directory Server from the Console
27
1.4.2. Starting and Stopping Directory Server from the Command Line
28
1.4.3. Starting and Stopping Admin Server
28
1.5. Starting the Console
29
1.5.1. Starting the Directory Server Console
29
1.5.2. Logging into Directory Server
30
1.5.3. Changing Login Identity
31
1.5.4. Viewing the Current Console Bind DN
31
1.6. Enabling LDAPI
32
1.7. Changing Directory Server Port Numbers
32
1.8. Creating a New Directory Server Instance
34
1.9. Setting the Directory Manager Information
35
1.10. Using Directory Server Plug-ins
36
1.10.1. Enabling Plug-ins
36
1.10.2. Setting the Plug-in Precedence
37
.Chapter
. . . . . . . . 2.
. . .Configuring
. . . . . . . . . . . . Directory
. . . . . . . . . . .Databases
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .39
...........
2.1. Creating and Maintaining Suffixes
39
2.1.1. Creating Suffixes
39
2.1.1.1. Creating a New Root Suffix Using the Console
41
2.1.1.2. Creating a New Sub Suffix Using the Console
42
2.1.1.3. Creating Root and Sub Suffixes from the Command Line
44
2.1.2. Maintaining Suffixes
47
2.1.2.1. Disabling a Suffix
47
2.1.2.2. Deleting a Suffix
48
2.2. Creating and Maintaining Databases
48
2.2.1. Creating Databases
48
2.2.1.1. Creating a New Database for an Existing Suffix Using the Console
50
2.2.1.2. Creating a New Database for a Single Suffix from the Command Line
51
2.2.1.3. Adding Multiple Databases for a Single Suffix
51
2.2.1.3.1. Adding the Custom Distribution Function to a Suffix Using the Directory Server
1
Red Hat D irectory Server 8.2 Administration Guide
Console
2.2.1.3.2. Adding the Custom Distribution Function to a Suffix Using the Command Line
2.2.2. Maintaining Directory Databases
2.2.2.1. Placing a Database in Read-Only Mode
2.2.2.1.1. Making a Database Read-Only Using the Console
2.2.2.1.2. Making a Database Read-Only from the Command Line
2.2.2.1.3. Placing the Entire Directory Server in Read-Only Mode
2.2.2.2. Deleting a Database
2.2.2.3. Configuring T ransaction Logs for Frequent Database Updates
2.2.3. Configuring Attribute Encryption
2.2.3.1. Encryption Keys
2.2.3.2. Encryption Ciphers
2.2.3.3. Configuring Attribute Encryption from the Console
2.2.3.4. Configuring Attribute Encryption Using the Command Line
2.2.3.5. Exporting and Importing an Encrypted Database
2.2.3.6. Updating Attribute Encryption Keys for New SSL/T LS Certificates
2.3. Creating and Maintaining Database Links
2.3.1. Creating a New Database Link
2.3.1.1. Creating a New Database Link Using the Console
2.3.1.2. Creating a Database Link from the Command Line
2.3.1.2.1. Providing Suffix Information
2.3.1.2.2. Providing Bind Credentials
2.3.1.2.3. Providing an LDAP URL
2.3.1.2.4. Providing a List of Failover Servers
2.3.1.2.5. Using Different Bind Mechanisms
2.3.1.2.6. Summary of Database Link Configuration Attributes
2.3.1.2.7. Database Link Configuration Example
2.3.2. Configuring the Chaining Policy
2.3.2.1. Chaining Component Operations
2.3.2.1.1. Chaining Component Operations Using the Console
2.3.2.1.2. Chaining Component Operations from the Command Line
2.3.2.2. Chaining LDAP Controls
2.3.2.2.1. Chaining LDAP Controls Using the Console
2.3.2.2.2. Chaining LDAP Controls from the Command Line
2.3.3. Maintaining Database Links
2.3.4. Configuring Database Link Defaults
2.3.5. Deleting Database Links
2.3.6. Database Links and Access Control Evaluation
2.4. Configuring Cascading Chaining
2.4.1. Overview of Cascading Chaining
2.4.2. Configuring Cascading Chaining Using the Console
2.4.3. Configuring Cascading Chaining from the Command Line
2.4.4. Detecting Loops
2.4.5. Summary of Cascading Chaining Configuration Attributes
2.4.6. Cascading Chaining Configuration Example
2.4.6.1. Configuring Server One
2.4.6.2. Configuring Server T wo
2.4.6.3. Configuring Server T hree
2.5. Using Referrals
2.5.1. Starting the Server in Referral Mode
2.5.2. Setting Default Referrals
2.5.2.1. Setting a Default Referral Using the Console
2.5.2.2. Setting a Default Referral from the Command Line
2.5.3. Creating Smart Referrals
2.5.3.1. Creating Smart Referrals Using the Directory Server Console
2
52
52
53
53
53
54
54
55
56
57
57
58
58
60
60
61
62
62
63
65
66
67
68
69
69
71
73
75
75
79
80
80
81
82
82
83
84
84
85
86
87
88
90
90
91
92
93
95
96
96
97
97
97
98
98
Table of Contents
2.5.3.2. Creating Smart Referrals from the Command Line
2.5.4. Creating Suffix Referrals
2.5.4.1. Creating Suffix Referrals Using the Console
2.5.4.2. Creating Suffix Referrals from the Command Line
100
101
101
102
.Chapter
. . . . . . . . 3.
. . .Creating
. . . . . . . . . Directory
. . . . . . . . . . Entries
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .104
.............
3.1. Managing Entries from the Directory Console
104
3.1.1. Creating a Root Entry
104
3.1.2. Creating Directory Entries
106
3.1.3. Modifying Directory Entries
108
3.1.3.1. Adding or Removing an Object Class to an Entry
109
3.1.3.2. Adding an Attribute to an Entry
111
3.1.3.3. Adding Very Large Attributes
113
3.1.3.4. Adding Attribute Values
113
3.1.3.5. Adding an Attribute Subtype
114
3.1.4. Deleting Directory Entries
115
3.2. Managing Entries from the Command Line
116
3.2.1. Providing Input from the Command Line
117
3.2.2. Creating a Root Entry from the Command Line
117
3.2.3. Adding Entries Using LDIF
118
3.2.4. Adding and Modifying Entries Using ldapmodify
118
3.2.4.1. Adding Entries Using ldapmodify
119
3.2.4.2. Modifying Entries Using ldapmodify
120
3.2.5. Deleting Entries Using ldapdelete
120
3.2.6. Using Special Characters
122
3.3. Using LDIF Update Statements to Create or Modify Entries
122
3.3.1. Adding an Entry Using LDIF
123
3.3.2. Renaming an Entry Using LDIF
125
3.3.2.1. A Note on Renaming Entries
125
3.3.3. Modifying an Entry Using LDIF
126
3.3.3.1. Adding Attributes to Existing Entries Using LDIF
126
3.3.3.2. Changing an Attribute Value Using LDIF
127
3.3.3.3. Deleting All Values of an Attribute Using LDIF
128
3.3.3.4. Deleting a Specific Attribute Value Using LDIF
128
3.3.4. Deleting an Entry Using LDIF
129
3.3.5. Modifying an Entry in an Internationalized Directory
129
3.4. T racking Modifications to Directory Entries
130
3.4.1. T racking Modifications to the Database through Update Sequence Numbers
130
3.4.1.1. Configuring the USN Plug-in
131
3.4.1.2. Cleaning up USN T ombstone Entries
131
3.4.2. T racking Entry Modifications through Operational Attributes
132
3.5. Maintaining Referential Integrity
134
3.5.1. How Referential Integrity Works
134
3.5.2. Using Referential Integrity with Replication
135
3.5.3. Enabling and Disabling Referential Integrity
135
3.5.3.1. Enabling and Disabling Referential Integrity in the Console
135
3.5.3.2. Enabling and Disabling Referential Integrity from the Command Line
137
3.5.4. Modifying the Update Interval
137
3.5.4.1. Modifying the Update Interval from the Console
137
3.5.4.2. Modifying the Update Interval from the Command Line
138
3.5.5. Modifying the Attribute List
138
3.5.5.1. Modifying the Attribute List from the Console
138
3.5.5.2. Modifying the Attribute List from the Command Line
139
.Chapter
........4
. ...Populating
. . . . . . . . . . . .Directory
. . . . . . . . . .Databases
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .14
. . .1. . . . . . . . . .
4.1. Importing Data
141
3
Red Hat D irectory Server 8.2 Administration Guide
4.1.1. Importing Entries with Large Attributes
4.1.2. Importing a Database from the Console
4.1.3. Initializing a Database from the Console
4.1.4. Importing from the Command Line
4.1.4.1. Importing Using the ldif2db Command-Line Script
4.1.4.2. Importing Using the ldif2db.pl Perl Script
4.1.4.3. Importing Using the ldif2ldap Command-Line Script
4.1.4.4. Importing through the cn=tasks Entry
4.2. Exporting Data
4.2.1. Exporting Directory Data to LDIF Using the Console
4.2.2. Exporting a Single Database to LDIF Using the Console
4.2.3. Exporting to LDIF from the Command Line
4.2.3.1. Exporting a Database Using db2ldif or db2ldif.pl
4.2.3.2. Exporting through the cn=tasks Entry
4.3. Backing up and Restoring Data
4.3.1. Backing up All Databases
4.3.1.1. Backing up All Databases from the Console
4.3.1.2. Backing up All Databases from the Command Line
4.3.1.3. Backing up the Database through the cn=tasks Entry
4.3.2. Backing up the dse.ldif Configuration File
4.3.3. Restoring All Databases
4.3.3.1. Restoring All Databases from the Console
4.3.3.2. Restoring Databases from the Command Line
4.3.3.2.1. Using the bak2db Command-Line Script
4.3.3.2.2. Using bak2db.pl Perl Script
4.3.3.2.3. Restoring the Database through the cn=tasks Entry
4.3.4. Restoring a Single Database
4.3.5. Restoring Databases T hat Include Replicated Entries
4.3.6. Restoring the dse.ldif Configuration File
141
141
143
144
144
145
146
147
147
148
149
150
150
152
153
153
153
154
155
155
156
156
157
157
157
158
159
159
160
.Chapter
. . . . . . . . 5.
. . .Managing
. . . . . . . . . . Attributes
. . . . . . . . . . . and
. . . . .Values
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
.............
5.1. Enforcing Attribute Uniqueness
161
5.1.1. Attribute Uniqueness Plug-in Syntax
161
5.1.2. Creating an Instance of the Attribute Uniqueness Plug-in
163
5.1.3. Configuring Attribute Uniqueness
163
5.1.3.1. Configuring Attribute Uniqueness Plug-ins from the Directory Server Console
163
5.1.3.2. Configuring Attribute Uniqueness Plug-ins from the Command Line
164
5.1.3.2.1. Specifying a Suffix or Subtree
164
5.1.3.2.2. Using the markerObjectClass and requiredObjectClass Keywords
164
5.1.4. Attribute Uniqueness Plug-in Syntax Examples
165
5.1.4.1. Specifying One Attribute and One Subtree
165
5.1.4.2. Specifying One Attribute and Multiple Subtrees
166
5.2. Assigning Class of Service
166
5.2.1. About the CoS Definition Entry
167
5.2.2. About the CoS T emplate Entry
167
5.2.3. How a Pointer CoS Works
167
5.2.4. How an Indirect CoS Works
168
5.2.5. How a Classic CoS Works
169
5.2.6. Handling Physical Attribute Values
169
5.2.7. Handling Multi-valued Attributes with CoS
170
5.2.8. Searches for CoS-Specified Attributes
171
5.2.9. Access Control and CoS
172
5.2.10. Managing CoS Using the Console
172
5.2.10.1. Creating a New CoS
172
5.2.10.2. Creating the CoS T emplate Entry
175
5.2.11. Managing CoS from the Command Line
181
4
Table of Contents
5.2.11.1. Creating the CoS Definition Entry from the Command Line
5.2.11.2. Creating the CoS T emplate Entry from the Command Line
5.2.11.3. Example of a Pointer CoS
5.2.11.4. Example of an Indirect CoS
5.2.11.5. Example of a Classic CoS
5.2.11.6. Searching for CoS Entries
5.2.12. Creating Role-Based Attributes
5.3. Linking Attributes to Manage Attribute Values
5.3.1. About Linking Attributes
5.3.2. Looking at the Linking Attributes Plug-in Syntax
5.3.3. Configuring Attribute Links
5.3.4. Cleaning up Attribute Links
5.3.4.1. Regenerating Linked Attributes Using fixup-linkedattrs.pl
5.3.4.2. Regenerating Linked Attributes Using ldapmodify
5.4. Assigning and Managing Unique Numeric Attribute Values
5.4.1. Looking at the DNA Plug-in Syntax
5.4.2. Configuring Unique Number Assignments
5.4.2.1. Configuring Unique Number Assignments
5.4.2.2. Editing the DNA Plug-in in the Console
181
182
183
183
184
185
185
187
187
188
189
190
190
191
191
193
196
196
197
.Chapter
. . . . . . . . 6.
. . .Managing
. . . . . . . . . . the
. . . . Directory
. . . . . . . . . . .Schema
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
.............
6.1. Overview of Schema
199
6.1.1. Default Schema Files
199
6.1.2. Object Classes
199
6.1.3. Attributes
200
6.1.4. Extending the Schema
201
6.1.5. Schema Replication
202
6.2. Managing Object Identifiers
202
6.3. Directory Server Attribute Syntaxes
202
6.4. Managing Custom Schema in the Console
206
6.4.1. Viewing Attributes and Object Classes
206
6.4.2. Creating Attributes
208
6.4.3. Creating Object Classes
209
6.4.4. Editing Custom Schema Elements
210
6.4.5. Deleting Schema
211
6.5. Managing Schema Using ldapmodify
212
6.5.1. Creating Attributes
212
6.5.2. Creating Object Classes
212
6.5.3. Deleting Schema
213
6.6. Creating Custom Schema Files
214
6.7. Dynamically Reloading Schema
215
6.7.1. Reloading Schema Using schema-reload.pl
215
6.7.2. Reloading Schema Using ldapmodify
216
6.7.3. Reloading Schema with Replication
217
6.7.4. Schema Reload Errors
217
6.8. T urning Schema Checking On and Off
217
6.9. Using Syntax Validation
219
6.9.1. About Syntax Validation
219
6.9.2. Syntax Validation and Other Directory Server Operations
219
6.9.3. Enabling or Disabling Syntax Validation
219
6.9.4. Enabling Strict Syntax Validation for DNs
220
6.9.5. Enabling Syntax Validation Warnings (Logging)
220
6.9.6. Validating the Syntax of Existing Attribute Values
221
.Chapter
. . . . . . . . 7.
. . .Managing
. . . . . . . . . . Indexes
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
.............
7.1. About Indexes
223
5
Red Hat D irectory Server 8.2 Administration Guide
7.1.1. About Index T ypes
7.1.2. About Default, System, and Standard Indexes
7.1.2.1. Overview of Default Indexes
7.1.2.2. Overview of System Indexes
7.1.2.3. Overview of Standard Indexes
7.1.3. Overview of the Searching Algorithm
7.1.4. Approximate Searches
7.1.5. Indexing Performance
7.1.5.1. Indexing Performance
7.1.5.2. Search Performance
7.1.5.3. Backwards Compatibility and Migration
7.1.6. Balancing the Benefits of Indexing
7.2. Creating Standard Indexes
7.2.1. Creating Indexes from the Server Console
7.2.2. Creating Indexes from the Command Line
7.3. Applying New Indexes to Existing Databases
7.3.1. Running the db2index.pl Script
7.3.2. Using a cn=tasks Entry to Create an Index
7.4. Creating Browsing (VLV) Indexes
7.4.1. Creating Browsing Indexes from the Server Console
7.4.2. Creating Browsing Indexes from the Command Line
7.4.2.1. Adding a Browsing Index Entry
7.4.2.2. Running the vlvindex Script
7.4.2.3. Using a cn=tasks Entry to Create a Browsing Index
7.4.3. Setting Access Control for VLV Information
7.5. Changing the Index Sort Order
7.5.1. Changing the Sort Order in the Console
7.5.2. Changing the Sort Order in the Command Line
7.6. Changing the Width for Indexed Substring Searches
7.7. Deleting Indexes
7.7.1. Deleting Indexes from the Server Console
7.7.2. Deleting Indexes from the Command Line
7.7.2.1. Deleting an Index Entry
7.7.2.2. Running the db2index.pl Script
7.7.3. Deleting Browsing Indexes from the Server Console
7.7.4. Deleting Browsing Indexes from the Command Line
7.7.4.1. Deleting a Browsing Index Entry
7.7.4.2. Running the vlvindex Script
223
224
224
226
227
227
228
229
229
230
230
231
232
232
234
235
236
236
237
237
238
239
240
241
241
242
242
242
243
244
244
245
245
246
247
248
248
249
.Chapter
. . . . . . . . 8.
. . .Finding
. . . . . . . .Directory
. . . . . . . . . . Entries
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
.............
8.1. Finding Entries Using the Directory Server Console
251
8.2. Using ldapsearch
252
8.2.1. ldapsearch Command-Line Format
253
8.2.2. Commonly Used ldapsearch Options
253
8.2.3. Using Special Characters
255
8.3. LDAP Search Filters
255
8.3.1. Using Attributes in Search Filters
256
8.3.2. Using Operators in Search Filters
256
8.3.3. Using Compound Search Filters
258
8.3.4. Using Matching Rules
259
8.4. Examples of Common ldapsearches
266
8.4.1. Returning All Entries
266
8.4.2. Specifying Search Filters on the Command Line
267
8.4.3. Searching the Root DSE Entry
267
8.4.4. Searching the Schema Entry
267
8.4.5. Using LDAP_BASEDN
267
6
Table of Contents
8.4.6. Displaying Subsets of Attributes
8.4.7. Searching for Operational Attributes
8.4.8. Specifying Search Filters Using a File
8.4.9. Specifying DNs T hat Contain Commas in Search Filters
8.4.10. Using Client Authentication When Searching
8.4.11. Searching with Specified Controls
8.4.12. Searching with Language Matching Rules
8.4.13. Searching for Attributes with Bit Field Values
8.5. Using Persistent Search
8.5.1. An Overview of Persistent Searches
8.5.2. Running a Persistent Search
8.6. Performing Dereferencing Searches
8.7. Using Simple Paged Results
267
268
268
268
269
269
269
269
270
271
272
272
274
.Chapter
. . . . . . . . 9.
. . .Managing
. . . . . . . . . . Replication
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
.............
9.1. Replication Overview
277
9.1.1. What Directory Units Are Replicated
277
9.1.2. Read-Write and Read-Only Replicas
277
9.1.3. Suppliers and Consumers
277
9.1.4. Changelog
278
9.1.5. Replication Identity
278
9.1.6. Replication Agreement
278
9.1.7. Replicating a Subset of Attributes with Fractional Replication
279
9.1.8. Compatibility with Earlier Versions of Directory Server
279
9.2. Replication Scenarios
279
9.2.1. Single-Master Replication
279
9.2.2. Multi-Master Replication
280
9.2.3. Cascading Replication
282
9.3. Creating the Supplier Bind DN Entry
283
9.4. Configuring Single-Master Replication
285
9.4.1. Configuring the Read-Write Replica on the Supplier Server
285
9.4.2. Configuring the Read-Only Replica on the Consumer
286
9.4.3. Creating the Replication Agreement
288
9.5. Configuring Multi-Master Replication
293
9.5.1. Configuring the Read-Write Replicas on the Supplier Servers
294
9.5.2. Configuring the Read-Only Replicas on the Consumer Servers
296
9.5.3. Setting up the Replication Agreements
298
9.5.4. Preventing Monopolization of the Consumer in Multi-Master Replication
303
9.6. Configuring Cascading Replication
304
9.6.1. Configuring the Read-Write Replica on the Supplier Server
305
9.6.2. Configuring the Read-Only Replica on the Consumer Server
306
9.6.3. Configuring the Read-Only Replica on the Hub
308
9.6.4. Setting up the Replication Agreements
311
9.7. Configuring Replication from the Command Line
316
9.7.1. Configuring Suppliers from the Command Line
317
9.7.2. Configuring Consumers from the Command Line
320
9.7.3. Configuring Hubs from the Command Line
320
9.7.4. Configuring Replication Agreements from the Command Line
321
9.7.5. Initializing Consumers Online from the Command Line
326
9.8. Making a Replica Updatable
327
9.9. Deleting the Changelog
328
9.9.1. Removing the Changelog
328
9.9.2. Moving the Changelog to a New Location
329
9.10. Initializing Consumers
329
9.10.1. When to Initialize a Consumer
330
9.10.2. Online Consumer Initialization Using the Console
330
7
Red Hat D irectory Server 8.2 Administration Guide
9.10.3. Initializing Consumers Online Using the Command Line
9.10.4. Manual Consumer Initialization Using the Command Line
9.10.4.1. Exporting a Replica to LDIF
9.10.4.2. Importing the LDIF File to the Consumer Server
9.10.5. Filesystem Replica Initialization
9.10.5.1. Initializing the Consumer Replica from the Backup Files
9.11. Forcing Replication Updates
9.11.1. Forcing Replication Updates from the Console
9.11.2. Forcing Replication Updates from the Command Line
9.12. Replication over SSL
9.13. Setting Replication T imeout Periods
9.14. Replicating o=NetscapeRoot for Admin Server Failover
9.15. Replication with Earlier Releases
9.15.1. Using Legacy Replication
9.15.2. Legacy Replication and Parent Object Classes
9.15.3. Configuring Legacy Replication
9.16. Using the Retro Changelog Plug-in
9.16.1. Enabling the Retro Changelog Plug-in
9.16.2. T rimming the Retro Changelog
9.16.3. Searching and Modifying the Retro Changelog
9.16.4. Retro Changelog and the Access Control Policy
9.17. Monitoring Replication Status
9.17.1. Monitoring Replication Status from the Directory Server Console
9.17.2. Monitoring Replication Status from Administration Express
9.18. Solving Common Replication Conflicts
9.18.1. Solving Naming Conflicts
9.18.1.1. Renaming an Entry with a Multi-Valued Naming Attribute
9.18.1.2. Renaming an Entry with a Single-Valued Naming Attribute
9.18.2. Solving Orphan Entry Conflicts
9.18.3. Solving Potential Interoperability Problems
9.19. T roubleshooting Replication-Related Problems
331
332
332
333
333
334
335
335
336
337
338
339
340
341
341
341
343
343
344
344
345
345
345
346
348
349
349
350
351
351
351
.Chapter
. . . . . . . . 10.
. . . . Synchronizing
. . . . . . . . . . . . . . . Red
. . . . .Hat
. . . .Directory
. . . . . . . . . . Server
. . . . . . . .with
. . . . Microsoft
. . . . . . . . . . .Active
. . . . . . .Directory
.........................
10.1. About Windows Sync
357 357
10.2. Configuring Windows Sync
359
10.2.1. Step 1: Configure SSL on Directory Server
360
10.2.2. Step 2: Configure the Active Directory Domain
362
10.2.3. Step 3: Select or Create the Sync Identity
364
10.2.4. Step 4: Install the Password Sync Service
365
10.2.5. Step 5: Configure the Password Sync Service
367
10.2.6. Step 6: Configure the Directory Server Database for Synchronization
368
10.2.6.1. Setting up the Directory Server from the Console
368
10.2.6.2. Setting up the Directory Server for Sync from the Command Line
370
10.2.7. Step 7: Create the Synchronization Agreement
371
10.2.7.1. Creating the Sync Agreement from the Console
371
10.2.7.2. Creating the Sync Agreement from the Command Line
373
10.2.8. Step 8: Configure Directory Server User and Group Entries for Synchronization
10.2.9. Step 9: Begin Synchronization
374 373
10.3. Synchronizing Users
374
10.3.1. User Attributes Synchronized between Directory Server and Active Directory
375
10.3.2. User Schema Differences between Red Hat Directory Server and Active Directory
10.3.2.1. Values for cn Attributes
376 376
10.3.2.2. Password Policies
377
10.3.2.3. Values for street and streetAddress
377
10.3.2.4. Constraints on the initials Attribute
377
10.3.3. Configuring User Sync for Directory Server Users
377
8
Table of Contents
10.3.3.1. Configuring User Sync in the Console
378
10.3.3.2. Configuring User Sync in the Command Line
379
10.3.4. Configuring User Sync for Active Directory Users
380
10.3.4.1. Configuring User Sync in the Console
380
10.3.4.2. Configuring User Sync in the Command Line
380
10.4. Synchronizing Groups
381
10.4.1. About Windows Group T ypes
381
10.4.2. Group Attributes Synchronized between Directory Server and Active Directory 382
10.4.3. Group Schema Differences between Red Hat Directory Server and Active Directory
10.4.4. Configuring Group Sync for Directory Server Groups
383 383
10.4.4.1. Configuring Group Sync in the Console
383
10.4.4.2. Configuring Group Sync in the Command Line
384
10.4.5. Configuring Group Sync for Active Directory Groups
385
10.4.5.1. Configuring Group Sync in the Console
385
10.4.5.2. Configuring Group Sync in the Command Line
386
10.5. Deleting and Resurrecting Entries
386
10.5.1. Deleting Entries
386
10.5.2. Resurrecting Entries
387
10.6. Sending Synchronization Updates
387
10.6.1. Performing a Manual Sync Update
387
10.6.2. Sending a T otal Update (Full Synchronization)
388
10.6.3. Sending Sync Updates in the Command Line
389
10.6.4. Checking Synchronization Status
389
10.7. Modifying the Sync Agreement
390
10.7.1. Editing the Sync Agreement in the Console
390
10.7.2. Adding and Editing the Sync Agreement in the Command Line
391
10.7.2.1. Creating a Basic Sync Agreement
391
10.7.2.2. Setting Sync Schedules
392
10.7.2.3. Changing Sync Connections
393
10.7.2.4. Sync Agreement Attributes
393
10.8. Configuring Unidirectional Synchronization
395
10.9. Managing the Password Sync Service
396
10.9.1. Modifying Password Sync
396
10.9.2. Starting and Stopping the Password Sync Service
396
10.9.3. Uninstalling Password Sync Service
398
10.9.4. Upgrading Password Sync
399
10.10. T roubleshooting
399
.Chapter
. . . . . . . . 11.
. . . . Organizing
. . . . . . . . . . . .Entries
. . . . . . . .with
. . . . .Groups,
. . . . . . . . .Roles,
. . . . . . and
. . . . .Views
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4. 01
............
11.1. Using Groups
401
11.1.1. Creating Static Groups in the Console
401
11.1.2. Creating Dynamic Groups in the Console
404
11.1.3. Creating Groups in the Command Line
407
11.1.4. Using the memberOf Attribute to Manage Group Membership Information
408
11.1.4.1. T he MemberOf Plug-in Syntax
409
11.1.4.2. Configuring an Instance of the MemberOf Plug-in from the Command Line
410
11.1.4.2.1. Editing the MemberOf Plug-in from the Console
410
11.1.4.2.2. Editing the MemberOf Plug-in from the Command Line
412
11.1.4.3. Synchronizing memberOf Values
412
11.1.4.3.1. Initializing and Regenerating memberOf Attributes Using fixup-memberof.pl
413
11.1.4.3.2. Initializing and Regenerating memberOf Attributes Using ldapmodify
413
11.2. Using Roles
413
11.2.1. About Roles
414
11.2.2. Creating a Managed Role
414
11.2.2.1. Creating a Managed Role in the Console
414
11.2.2.2. Creating Managed Roles through the Command Line
416
9
Red Hat D irectory Server 8.2 Administration Guide
11.2.3. Creating a Filtered Role
11.2.3.1. Creating a Filtered Role in the Console
11.2.3.2. Creating a Filtered Role through the Command Line
11.2.4. Creating a Nested Role
11.2.4.1. Creating a Nested Role in the Console
11.2.4.2. Creating Nested Role through the Command Line
11.2.5. Editing and Assigning Roles to an Entry
11.2.6. Viewing Roles for an Entry through the Command Line
11.2.7. Making a Role Inactive or Active
11.2.8. Viewing the Activation Status for Entries
11.2.9. About Deleting Roles
11.2.10. Using Roles Securely
11.3. Using Views
11.3.1. Creating Views in the Console
11.3.2. Creating Views from the Command Line
417
417
419
420
420
422
423
424
425
425
426
426
427
428
431
.Chapter
. . . . . . . . 12.
. . . . Managing
. . . . . . . . . . .Access
. . . . . . . .Control
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. .33
...........
12.1. Access Control Principles
433
12.1.1. ACI Structure
433
12.1.2. ACI Placement
433
12.1.3. ACI Evaluation
434
12.1.4. ACI Limitations
434
12.2. Default ACIs
435
12.3. Creating ACIs Manually
435
12.3.1. T he ACI Syntax
436
12.3.2. Defining T argets
436
12.3.2.1. T argeting a Directory Entry
438
12.3.2.2. T argeting Attributes
439
12.3.2.3. T argeting Both an Entry and Attributes
439
12.3.2.4. T argeting Entries or Attributes Using LDAP Filters
440
12.3.2.5. T argeting Attribute Values Using LDAP Filters
440
12.3.2.6. T argeting a Single Directory Entry
441
12.3.3. Defining Permissions
442
12.3.3.1. Allowing or Denying Access
442
12.3.3.2. Assigning Rights
442
12.3.3.3. Rights Required for LDAP Operations
444
12.3.3.4. Permissions Syntax
445
12.3.3.5. Access Control and the modrdn Operation
445
12.4. Bind Rules
445
12.4.1. Bind Rule Syntax
446
12.4.2. Defining User Access - userdn Keyword
447
12.4.2.1. Anonymous Access (anyone Keyword)
448
12.4.2.2. General Access (all Keyword)
448
12.4.2.3. Self Access (self Keyword)
448
12.4.2.4. Parent Access (parent Keyword)
448
12.4.2.5. LDAP URLs
448
12.4.2.6. Wildcards
449
12.4.2.7. Examples
449
12.4.3. Defining Group Access - groupdn Keyword
455
12.4.4. Defining Role Access - roledn Keyword
459
12.4.5. Defining Access Based on Value Matching
459
12.4.5.1. Using the userattr Keyword
460
12.4.5.1.1. Example with USERDN Bind T ype
460
12.4.5.1.2. Example with GROUPDN Bind T ype
460
12.4.5.1.3. Example with ROLEDN Bind T ype
461
12.4.5.1.4. Example with LDAPURL Bind T ype
461
10
Table of Contents
12.4.5.1.5. Example with Any Attribute Value
461
12.4.5.1.6. Using the userattr Keyword with Inheritance
462
12.4.5.1.7. Granting Add Permission Using the userattr Keyword
463
12.4.6. Defining Access from a Specific IP Address
464
12.4.7. Defining Access from a Specific Domain
464
12.4.8. Requiring a Certain Level of Security in Connections
465
12.4.9. Defining Access at a Specific T ime of Day or Day of Week
465
12.4.9.1. Examples
466
12.4.10. Defining Access Based on Authentication Method
467
12.4.10.1. Examples
467
12.4.11. Using Boolean Bind Rules
468
12.5. Creating ACIs from the Console
468
12.5.1. Displaying the Access Control Editor
469
12.5.2. Creating a New ACI
470
12.5.3. Editing an ACI
474
12.5.4. Deleting an ACI
475
12.6. Viewing ACIs
476
12.7. Checking Access Rights on Entries (Get Effective Rights)
477
12.7.1. Rights Shown with a Get Effective Rights Search
477
12.7.2. T he Format of a Get Effective Rights Search
478
12.7.2.1. General Examples on Checking Access Rights
479
12.7.2.2. Examples of Get Effective Rights Searches for Non-Existent Attributes
482
12.7.2.3. Examples of Get Effective Rights Searches for Specific Attributes or Object Classes
12.7.2.4. Examples of Get Effective Rights Searches for Operational Attributes
484 483
12.7.2.5. Examples of Get Effective Rights Results and Access Control Rules
485
12.7.3. Using Get Effective Rights from the Console
486
12.7.4. Get Effective Rights Return Codes
487
12.8. Logging Access Control Information
488
12.9. Access Control Usage Examples
489
12.9.1. Granting Anonymous Access
490
12.9.1.1. ACI "Anonymous example.com"
490
12.9.1.2. ACI "Anonymous World"
491
12.9.2. Granting Write Access to Personal Entries
492
12.9.2.1. ACI "Write example.com"
492
12.9.2.2. ACI "Write Subscribers"
495
12.9.3. Restricting Access to Key Roles
497
12.9.3.1. ACI "Roles"
498
12.9.4. Granting a Group Full Access to a Suffix
498
12.9.4.1. ACI "HR"
499
12.9.5. Granting Rights to Add and Delete Group Entries
501
12.9.5.1. ACI "Create Group"
501
12.9.5.2. ACI "Delete Group"
503
12.9.6. Granting Conditional Access to a Group or Role
503
12.9.6.1. ACI "HostedCompany1"
504
12.9.7. Denying Access
506
12.9.7.1. ACI "Billing Info Read"
506
12.9.7.2. ACI "Billing Info Deny"
507
12.9.8. Setting a T arget Using Filtering
508
12.9.9. Allowing Users to Add or Remove T hemselves from a Group
508
12.9.9.1. ACI "Group Members"
509
12.9.10. Setting an ACI to Require a Certain Security Strength Factor for Some Operations
510
12.9.11. Defining Permissions for DNs T hat Contain a Comma
510
12.9.12. Proxied Authorization ACI Example
511
12.10. Advanced Access Control: Using Macro ACIs
512
12.10.1. Macro ACI Example
512
11
Red Hat D irectory Server 8.2 Administration Guide
12.10.2. Macro ACI Syntax
12.10.2.1. Macro Matching for ($dn)
12.10.2.2. Macro Matching for [$dn]
12.10.2.3. Macro Matching for ($attr.attrName)
12.11. Access Control and Replication
12.12. Compatibility with Earlier Releases
513
514
515
516
516
516
.Chapter
. . . . . . . . 13.
. . . .Managing
. . . . . . . . . . .User
. . . . .Authentication
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 518
.............
13.1. Managing the Password Policy
518
13.1.1. Configuring the Password Policy
518
13.1.1.1. Configuring a Global Password Policy Using the Console
519
13.1.1.2. Configuring a Subtree/User Password Policy Using the Console
521
13.1.1.3. Configuring a Global Password Policy Using the Command Line
523
13.1.1.4. Configuring Subtree/User Password Policy Using the Command Line
529
13.1.1.5. Manually Setting Default Password Syntax Checking for Local Password Policies 531
13.1.2. Setting User Passwords
532
13.1.3. Password Change Extended Operation
532
13.2. Configuring the Account Lockout Policy
534
13.2.1. Configuring the Account Lockout Policy Using the Console
534
13.2.2. Configuring the Account Lockout Policy Using the Command Line
535
13.2.3. Replicating Account Lockout Attributes
537
13.2.3.1. Managing the Account Lockouts and Replication
537
13.2.3.2. Configuring Directory Server to Replicate Password Policy Attributes
538
13.2.3.3. Configuring Fractional Replication for Password Policy Attributes
538
13.3. Synchronizing Passwords
539
13.4. Setting Resource Limits Based on the Bind DN
540
13.4.1. Setting Resource Limits Using the Console
540
13.4.2. Setting Resource Limits Using the Command Line
541
13.4.3. Setting Resource Limits on Anonymous Binds
542
13.5. Enabling Different T ypes of Binds
542
13.5.1. Requiring Secure Binds
543
13.5.2. Disabling Anonymous Binds
543
13.5.3. Allowing Unauthenticated Binds
544
13.5.4. Configuring Autobind
545
13.5.4.1. Overview of Autobind and LDAPI
545
13.5.4.2. Configuring Autobind
546
13.6. Using Pass-through Authentication
547
13.6.1. PT A Plug-in Syntax
549
13.6.2. Configuring the PT A Plug-in
551
13.6.2.1. Configuring the Servers to Use a Secure Connection
552
13.6.2.2. Specifying the Authenticating Directory Server
552
13.6.2.3. Specifying the Pass-through Subtree
553
13.6.2.4. Configuring the Optional Parameters
553
13.6.3. PT A Plug-in Syntax Examples
554
13.6.3.1. Specifying One Authenticating Directory Server and One Subtree
554
13.6.3.2. Specifying Multiple Authenticating Directory Servers
554
13.6.3.3. Specifying One Authenticating Directory Server and Multiple Subtrees
555
13.6.3.4. Using Non-Default Parameter Values
555
13.6.3.5. Specifying Different Optional Parameters and Subtrees for Different Authenticating
Directory Servers
555
13.7. Using PAM for Pass-through Authentication
556
13.7.1. PAM Pass-through Authentication Configuration Options
556
13.7.1.1. Specifying the Subtrees to T arget for PAM PT A
557
13.7.1.2. Setting PAM PT A Mappings
558
13.7.1.3. Configuring General PAM PT A Settings
559
13.7.2. Configuring PAM Pass-through Authentication
559
12
Table of Contents
13.8. Inactivating Users and Roles
13.8.1. Activating and Inactivating Users and Roles Using the Console
13.8.2. Viewing Inactive Users and Roles
13.8.3. Inactivating and Activating Users and Roles Using the Command Line
560
560
561
561
.Chapter
. . . . . . . . 14
. . . .. Configuring
. . . . . . . . . . . . .Secure
. . . . . . . .Connections
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 563
.............
14.1. Requiring Secure Connections
563
14.2. Using T LS/SSL
564
14.2.1. Enabling T LS/SSL: Summary of Steps
564
14.2.2. Obtaining and Installing Server Certificates
565
14.2.2.1. Generating a Certificate Request
565
14.2.2.2. Sending a Certificate Request
567
14.2.2.3. Installing the Certificate
568
14.2.2.4. T rusting the Certificate Authority
570
14.2.2.5. Confirming T hat T he New Certificates Are Installed
571
14.2.3. Configuring the Directory Server to Run in SSL/T LS
571
14.2.3.1. Enabling T LS/SSL Only in the Directory Server
572
14.2.3.2. Enabling T LS/SSL in the Directory Server, Admin Server, and Console
575
14.2.3.3. Creating a Password File for the Directory Server
578
14.2.3.4. Creating a Password File for the Admin Server
578
14.2.4. Command-Line Functions for Start T LS
579
14.2.5. Using certutil
580
14.2.5.1. Creating Directory Server Certificates through the Command Line
581
14.2.5.2. certutil Usage
583
14.2.6. Managing Certificates Used by the Directory Server Console
585
14.2.7. Updating Attribute Encryption for New SSL/T LS Certificates
586
14.2.8. Using External Security Devices
586
14.2.8.1. Installing PKCS#11 Modules T hrough the Directory Server Console
587
14.2.8.2. Installing PKCS#11 Modules T hrough the Command Line
587
14.2.9. Setting Security Preferences
588
14.2.9.1. Available Ciphers
588
14.2.9.2. Selecting the Encryption Cipher
590
14.2.10. Using Certificate-Based Authentication
591
14.2.10.1. Configuring Directory Server to Accept Certificate-Based Authentication from LDAP
Clients
593
14.2.10.2. Mapping DNs to Certificates
594
14.2.10.3. Editing the certmap.conf File
597
14.2.10.4. Example certmap.conf Mappings
598
14.2.10.5. Allowing and Requiring Client Authentication to the Console
599
14.2.10.6. Connecting to the Directory Server with Certificate-Based Authentication
600
14.2.10.7. Forcing SASL/EXT ERNAL Mechanisms for BIND Requests
600
14.2.11. Managing Certificates for the Directory Server
601
14.2.11.1. Renewing Certificates
601
14.2.11.2. Changing the CA T rust Options
602
14.2.11.3. Changing Security Device Passwords
603
14.2.11.4. Adding Certificate Revocation Lists
604
14.3. Using SASL
605
14.3.1. About SASL Identity Mapping
606
14.3.2. Default SASL Mappings for Directory Server
609
14.3.3. Authentication Mechanisms for SASL in Directory Server
610
14.3.4. About Kerberos with Directory Server
610
14.3.4.1. About Principals and Realms
611
14.3.4.2. About the KDC Server and Keytabs
611
14.3.4.3. Setting up Kerberos on Solaris
613
14.3.5. Configuring SASL Identity Mapping
614
14.3.5.1. Configuring SASL Identity Mapping from the Console
614
13
Red Hat D irectory Server 8.2 Administration Guide
14.3.5.2. Configuring SASL Identity Mapping from the Command Line
14.3.6. Configuring SASL Authentication at Directory Server Startup
14.3.7. Using an External Keytab
615
615
616
.Chapter
. . . . . . . . 15.
. . . . Monitoring
. . . . . . . . . . . .Server
. . . . . . . and
. . . . .Database
. . . . . . . . . . Activity
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 617
.............
15.1. T ypes of Directory Server Log Files
617
15.2. Viewing Log Files
617
15.3. Configuring Logs
618
15.3.1. Enabling or Disabling Logs
618
15.3.2. Defining a Log File Rotation Policy
619
15.3.3. Defining a Log File Deletion Policy
620
15.3.4. Manual Log File Rotation
621
15.3.5. Configuring Log Levels
621
15.4. Replacing Log Files with a Named Pipe
623
15.4.1. Using the Named Pipe for Logging
623
15.4.2. Starting the Named Pipe with the Server
624
15.4.3. Using Plug-ins with the Named Pipe Log
625
15.4.3.1. Loading Plug-ins with the Named Pipe Log Script
625
15.4.3.2. Writing Plug-ins to Use with the Named Pipe Log Script
626
15.5. Monitoring Server Activity
627
15.5.1. Monitoring the Server from the Directory Server Console
627
15.5.2. Monitoring the Directory Server from the Command Line
632
15.6. Monitoring Database Activity
634
15.6.1. Monitoring Database Activity from the Directory Server Console
634
15.6.2. Monitoring Databases from the Command Line
639
15.7. Monitoring Database Link Activity
642
15.8. Enabling and Disabling Counters
642
.Chapter
. . . . . . . . 16.
. . . . Monitoring
. . . . . . . . . . . .Directory
. . . . . . . . . .Server
. . . . . . . Using
. . . . . . .SNMP
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .64
. . .4. . . . . . . . . .
16.1. About SNMP
644
16.2. Configuring the Master Agent
644
16.3. Configuring the Subagent
645
16.3.1. Creating the Subagent Configuration File
645
16.3.1.1. agentx-master
645
16.3.1.2. agent-logdir
645
16.3.1.3. server
646
16.3.2. Starting the Subagent
646
16.3.3. T esting the Subagent
646
16.4. Configuring SNMP T raps
647
16.5. Configuring the Directory Server for SNMP
647
16.6. Using the Management Information Base
648
16.6.1. Operations T able
648
16.6.2. Entries T able
651
16.6.3. Entity T able
651
16.6.4. Interaction T able
652
.Chapter
. . . . . . . . 17.
. . . . Planning
. . . . . . . . . .for
. . . Disaster
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .654
.............
17.1. Identifying Potential Scenarios
654
17.2. Defining the T ype of Rollover
655
17.3. Identifying Useful Directory Server Features for Disaster Recovery
655
17.3.1. Multi-Master Replication for Disaster Recovery
655
17.3.2. Chaining Databases for Disaster Recovery
656
17.3.3. Backing up Directory Data for Disaster Recovery
656
17.3.4. Using a Named Pipe Script for Disaster Recovery
657
17.4. Defining the Recovery Process
657
17.5. Basic Example: Performing a Recovery
657
14
Table of Contents
.LDAP
. . . . . .Data
. . . . . Interchange
. . . . . . . . . . . . . Format
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 659
.............
A.1. About the LDIF File Format
659
A.2. Continuing Lines in LDIF
660
A.3. Representing Binary Data
661
A.3.1. Standard LDIF Notation
661
A.3.2. Base-64 Encoding
661
A.4. Specifying Directory Entries Using LDIF
662
A.4.1. Specifying Domain Entries
662
A.4.2. Specifying Organizational Unit Entries
663
A.4.3. Specifying Organizational Person Entries
664
A.5. Defining Directories Using LDIF
665
A.5.1. LDIF File Example
666
A.6. Storing Information in Multiple Languages
668
.LDAP
. . . . . .URLs
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 670
.............
B.1. Components of an LDAP URL
670
B.2. Escaping Unsafe Characters
672
B.3. Examples of LDAP URLs
672
.Internationalization
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 675
.............
C.1. About Locales
675
C.2. Supported Locales
675
C.3. Supported Language Subtypes
678
C.4. Searching an Internationalized Directory
679
C.4.1. Matching Rule Formats
680
C.4.1.1. Using an OID for the Matching Rule
680
C.4.1.2. Using a Language T ag for the Matching Rule
681
C.4.1.3. Using an OID and Suffix for the Matching Rule
681
C.4.1.4. Using a Language T ag and Suffix for the Matching Rule
681
C.4.2. Supported Search T ypes
681
C.4.3. International Search Examples
682
C.4.3.1. Less-T han Example
682
C.4.3.2. Less-T han or Equal-to Example
682
C.4.3.3. Equality Example
683
C.4.3.4. Greater-T han or Equal-to Example
683
C.4.3.5. Greater-T han Example
683
C.4.3.6. Substring Example
684
C.5. T roubleshooting Matching Rules
684
.Glossary
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 685
.............
A
685
B
686
C
687
D
689
E
691
F
691
G
691
H
692
I
692
K
693
L
693
M
694
N
696
O
696
P
697
R
698
15
Red Hat D irectory Server 8.2 Administration Guide
S
T
U
V
X
700
703
704
704
704
.Index
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .704
.............
A
704
B
710
C
712
D
717
E
722
F
723
G
724
H
725
I
725
J
727
K
727
L
727
M
732
N
733
O
734
P
735
Q
739
R
739
S
742
T
749
U
750
V
750
W
751
16
Table of Contents
17
Red Hat D irectory Server 8.2 Administration Guide
Preface
Red Hat Directory Server (Directory Server) is a powerful and scalable distributed directory server
based on the industry-standard Lightweight Directory Access Protocol (LDAP). Directory Server is the
cornerstone for building a centralized and distributed data repository that can be used in your intranet,
over your extranet with your trading partners, or over the public Internet to reach your customers.
T his Administrator's Guide describes all of the administration tasks you need to perform to maintain
Directory Server.
1. Directory Server Overview
Directory Server provides the following key features:
Multi-master replication — Provides a highly available directory service for both read and write
operations. Multi-master replication can be combined with simple and cascading replication scenarios
to provide a highly flexible and scalable replication environment.
Chaining and referrals — Increases the power of your directory by storing a complete logical view of
your directory on a single server while maintaining data on a large number of Directory Servers
transparently for clients.
Roles and classes of service — Provides a flexible mechanism for grouping and sharing attributes
between entries in a dynamic fashion.
Improved access control mechanisms — Provides support for macros that dramatically reduce the
number of access control statements used in the directory and increase the scalability of access
control evaluation.
Resource-limits by bind DN — Grants the power to control the amount of server resources allocated
to search operations based on the bind DN of the client.
Multiple databases — Provides a simple way of breaking down your directory data to simplify the
implementation of replication and chaining in your directory service.
Password policy and account lockout — Defines a set of rules that govern how passwords and user
accounts are managed in the Directory Server.
T LS and SSL — Provides secure authentication and communication over the network, using the
Mozilla Network Security Services (NSS) libraries for cryptography.
T he major components of Directory Server include the following:
An LDAP server — T he LDAP v3-compliant network daemon.
Directory Server Console — A graphical management console that dramatically reduces the effort of
setting up and maintaining your directory service.
SNMP agent — Can monitor the Directory Server using the Simple Network Management Protocol
(SNMP).
2. Examples and Formatting
Each of the examples used in this guide, such as file locations and commands, have certain defined
conventions.
2.1. Command and File Examples
All of the examples for Red Hat Directory Server commands, file locations, and other usage are given for
Red Hat Enterprise Linux 5 (64-bit) systems. Be certain to use the appropriate commands and files for
your platform.
18
Preface
Example 1. Example Command
T o start the Red Hat Directory Server:
service dirsrv start
2.2. Tool Locations
T he tools for Red Hat Directory Server are located in the /usr/bin and the /usr/sbin directories.
T hese tools can be run from any location without specifying the tool location.
2.3. LDAP Locations
T here is an important consideration with the Red Hat Directory Server tools. T he LDAP tools referenced
in this guide are Mozilla LDAP, installed with Red Hat Directory Server in the /usr/lib64 /m ozldap
directory on Red Hat Enterprise Linux 5 (64-bit) (or /usr/lib/m ozldap for Red Hat Enterprise Linux 5
(32-bit) systems).
However, Red Hat Enterprise Linux systems also include LDAP tools from OpenLDAP in the /usr/bin
directory. It is possible to use the OpenLDAP commands as shown in the examples, but you must use
the -x argument to disable SASL, which OpenLDAP tools use by default.
2.4. Text Formatting and Styles
Certain words are represented in different fonts, styles, and weights. Different character formatting is
used to indicate the function or purpose of the phrase being highlighted.
Formatting Style
Purpose
Monospace font
Monospace is used for commands, package
names, files and directory paths, and any text
displayed in a prompt.
Monospace
with a
background
T his type of formatting is used for anything
entered or returned in a command prompt.
Italicized text
Any text which is italicized is a variable, such as
instance_name or hostname. Occasionally, this is
also used to emphasize a new term or other
phrase.
Bolded text
Most phrases which are in bold are application
names, such as Cygwin, or are fields or options
in a user interface, such as a User Nam e
Here: field or Save button.
Other formatting styles draw attention to important text.
19
Red Hat D irectory Server 8.2 Administration Guide
NOTE
A note provides additional information that can help illustrate the behavior of the system or
provide more detail for a specific issue.
IMPORTANT
Important information is necessary, but possibly unexpected, such as a configuration change that
will not persist after a reboot.
WARNING
A warning indicates potential data loss, as may happen when tuning hardware for maximum
performance.
3. Additional Reading
T he Directory Server Administrator's Guide describes how to set up, configure, and administer Red Hat
Directory Server and its contents. this manual does not describe many of the basic directory and
architectural concepts that you need to deploy, install, and administer a directory service successfully.
T hose concepts are contained in the Red Hat Directory Server Deployment Guide. You should read that
book before continuing with this manual.
When you are familiar with Directory Server concepts and have done some preliminary planning for your
directory service, install the Directory Server. T he instructions for installing the various Directory Server
components are contained in the Red Hat Directory Server Installation Guide. Many of the scripts and
commands used to install and administer the Directory Server are explained in detail in the Red Hat
Directory Server Configuration and Command-Line Tool Reference.
Also, Managing Servers with Red Hat Console contains general background information on how to use
the Red Hat Console. You should read and understand the concepts in that book before you attempt to
administer Directory Server.
T he document set for Directory Server contains the following guides:
Red Hat Directory Server Release Notes contain important information on new features, fixed bugs,
known issues and workarounds, and other important deployment information for this specific version
of Directory Server.
Red Hat Directory Server Deployment Guide provides an overview for planning a deployment of the
Directory Server.
Red Hat Directory Server Administrator's Guide contains procedures for the day-to-day maintenance
of the directory service. Includes information on configuring server-side plug-ins.
Red Hat Directory Server Configuration and Command-Line Tool Reference provides reference
information on the command-line scripts, configuration attributes, and log files shipped with Directory
Server.
Red Hat Directory Server Installation Guide contains procedures for installing your Directory Server
as well as procedures for migrating from a previous installation of Directory Server.
20
Preface
Red Hat Directory Server Schema Reference provides reference information about the Directory
Server schema.
Red Hat Directory Server Plug-in Programmer's Guide describes how to write server plug-ins in
order to customize and extend the capabilities of Directory Server.
Using Red Hat Console gives an overview of the primary user interface and how it interacts with the
Directory Server and Admin Server, as well as how to perform basic management tasks through the
main Console window.
Using the Admin Server describes the different tasks and tools associated with the Admin Server
and how to use the Admin Server with the Configuration and User Directory Server instances.
For the latest information about Directory Server, including current release notes, complete product
documentation, technical notes, and deployment information, see the Red Hat Directory Server
documentation site at http://www.redhat.com/docs/manuals/dir-server/.
4. Giving Feedback
If there is any error in this Administrator's Guide or there is any way to improve the documentation,
please let us know. Bugs can be filed against the documentation for Red Hat Directory Server through
Bugzilla, http://bugzilla.redhat.com/bugzilla. Make the bug report as specific as possible, so we can be
more effective in correcting any issues:
1. Select the Red Hat Directory Server product.
2. Set the component to Doc - adm inistration-guide.
3. Set the version number to 8.2.
4. For errors, give the page number (for the PDF) or URL (for the HT ML), and give a succinct
description of the problem, such as incorrect procedure or typo.
For enhancements, put in what information needs to be added and why.
5. Give a clear title for the bug. For example, "Incorrect com m and exam ple for setup
script options" is better than "Bad exam ple".
We appreciate receiving any feedback — requests for new sections, corrections, improvements,
enhancements, even new ways of delivering the documentation or new styles of docs. You are welcome
to contact Red Hat Content Services directly at [email protected].
5. Documentation History
Revision 8.2.8-3.4 00
Rebuild with publican 4.0.0
2013-10-31
Rüdiger Landmann
Revision 8.2.8-3
Rebuild for Publican 3.0
2012-07-18
Anthony T owns
Revision 8.2-15
April 2, 2012
Removing reference to IPv6 suport for ACIs.
Ella Deon Lackey
Revision 8.2-14
January 31, 2012
Ella Deon Lackey
Adding an example for the new nested DN validation attribute, Errata RHBA-2012:0064.
Revision 8.2-13
July 1, 2011
Backing out some attribute syntax changes, Bugzilla 524991.
Ella Deon Lackey
21
Red Hat D irectory Server 8.2 Administration Guide
Revision 8.2-12
June 3, 2011
Ella Deon Lackey
Adding a note about nsrolefilter not supporting virtual attributes, Bugzilla 329751.
Revision 8.2-11
June 3, 2011
Fixing ldapmodify proxy account example, Bugzilla 707585.
Ella Deon Lackey
Revision 8.2-10
May 23, 2011
Fixing a log setting example, Bugzilla 703711.
Ella Deon Lackey
Revision 8.2-9
April 4 , 2011
Persistent search review, Bugzilla 367741.
Replication chapter typos, Bugzilla 690637 and Bugzilla 690823.
Ella Deon Lackey
Revision 8.2-8
Fixing typos, Bugzilla 680150.
February 24 , 2011
Ella Deon Lackey
Revision 8.2-7
Fixing typos, Bugzilla 662860.
January 6, 2011
Ella Deon Lackey
Revision 8.2-6
November 19, 2010
Ella Deon Lackey
Added warning note for configuring SSL in the Console, per Bugzilla #488951.
Added information on loading certificates in the Console security databases, per Bugzilla #632420.
Revision 8.2-5
November 15, 2010
Lee Carlon
Added information regarding replication of memberOf attribute in multi-master and single suppier, per
Bugzilla #536865.
Added information regarding requirement for inetUser and memberOf attribute for MemberOf plug-in, per
Bugzilla #642776.
Updated screen short for connection properties for windows sync, per Bugzilla #577400/
Updated the name of a reference guide, per Bugzilla #578642.
Revision 8.2-4
November 1, 2010
Ella Deon Lackey
Corrected example for using the markerobjectclass keywork with Attribute Uniqueness, per Bugzilla
#646651.
Revision 8.2-3
October 12, 2010
Ella Deon Lackey
Corrected example for enabling the Referential Integriry plug-in, per Bugzilla #640876.
Revision 8.2-2
September 17, 2010
Ella Deon Lackey
Clarified some comments in the account lockout section from a customer request, per Bugzilla #621646.
Adding section on using modutil to load PKCS#11 modules, per Bugzilla #629411.
Clarifying the backup procedures with db2bak and db2bak.pl, per Bugzilla #632014.
Revision 8.2-1
August 10, 2010
Ella Deon Lackey
Fixing user account management example, per Bugzilla #622525.
Adding note to DNA section that the DNA plug-in cannot check multiple backends to allocate numbers,
per Bugzilla #621727.
Revision 8.2-0
22
August 3, 2010
Ella Deon Lackey
Preface
Initial draft for Red Hat Directory Server version 8.2.
23
Red Hat D irectory Server 8.2 Administration Guide
Chapter 1. Basic Red Hat Directory Server Settings
Red Hat Directory Server product includes a directory service, an administration server to manage
multiple server instances, and a Java-based console to manage server instances through a graphical
interface. T his chapter provides an overview of the basic tasks for administering a directory service.
T he Directory Server is a robust, scalable server designed to manage an enterprise-wide directory of
users and resources. It is based on an open-systems server protocol called the Lightweight Directory
Access Protocol (LDAP). Directory Server runs the ns-slapd daemon on the host machine. T he server
manages the directory databases and responds to client requests.
Directory Server 8.2 is comprised of several components, which work in tandem:
T he Directory Server is the core LDAP server daemon. It is compliant with LDAP v3 standards. T his
component includes command-line server management and administration programs and scripts for
common operations like export and backing up databases.
T he Directory Server Console is the user interface that simplifies managing users, groups, and other
LDAP data for your enterprise. T he Console is used for all aspects of server management, including
making backups; configuring security, replication, and databases; adding entries; and monitoring
servers and viewing statistics.
T he Admin Server is the management agent which administers Directory Server instances. It
communicates with the Directory Server Console and performs operations on the Directory Server
instances. It also provides a simple HT ML interface and online help pages.
Most Directory Server administrative tasks are available through the Directory Server Console, but it is
also possible to administer the Directory Server by manually editing the configuration files or by using
command-line utilities.
1.1. System Requirements
T his section contains information related to installing and upgrading Red Hat Directory Server 8.2,
including prerequisites and hardware or platform requirements.
1.1.1. Required JDK
Red Hat Directory Server 8.2 requires Sun JRE 1.6.0 or OpenJDK 1.6.0 for Red Hat Enterprise Linux 4
and 5.
IMPORTANT
When the new JDK is installed for Directory Server 8.2, it is no longer possible to manage older
instances of Directory Server using the Directory Server Console because the required JDKs for
the different Directory Server versions are different. You must migrate any older instance to
Directory Server 8.2 if you need to manage that instance with the Directory Server Console.
1.1.2. Directory Server Supported Platforms
Directory Server 8.2 is supported on the following platforms:
Red Hat Enterprise Linux 4 i386 (32-bit)
Red Hat Enterprise Linux 4 x86_64 (64-bit)
Red Hat Enterprise Linux 5 i386 (32-bit)
24
Chapter 1. Basic Red Hat D irectory Server Settings
Red Hat Enterprise Linux 5 x86_64 (64-bit)
NOTE
Red Hat Directory Server 8.2 is supported running on a virtual guest on a Red Hat Enterprise
Linux virtual server.
Solaris 9 SPARC (64-bit)
1.1.3. Directory Server Console Supported Platforms
T he Directory Server Console is supported on the following platforms:
Red Hat Enterprise Linux 4 i386 (32-bit)
Red Hat Enterprise Linux 4 x86_64 (64-bit)
Red Hat Enterprise Linux 5 i386 (32-bit)
Red Hat Enterprise Linux 5 x86_64 (64-bit)
Solaris 9 SPARC (64-bit)
Windows XP Professional
Windows Server 2003
Windows Server 2008 (32-bit)
Windows Server 2008 (64-bit)
1.1.4. Password Sync Service Platforms
T he Password Sync Service runs on these Windows platforms:
Windows 2003 Active Directory (32-bit)
Windows 2003 Active Directory (64-bit)
Windows 2008 Active Directory (32-bit)
Windows 2008 Active Directory (64-bit)
1.1.5. Web Application Browser Support
Directory Server 8.2 supports the following browsers to access web-based interfaces, such as Admin
Express and online help tools:
Firefox 3.x
Microsoft Internet Explorer 6.0 and higher
1.2. Directory Server File Locations
Red Hat Directory Server 8.2 conforms to the Filesystem Hierarchy Standards. For more information on
FHS, see the FHS homepage, http://www.pathname.com/fhs/. T he files and directories installed with
Directory Server are listed in the tables below for each supported platform.
In the file locations listed in the following tables, instance is the server instance name that was given
during setup. By default, this is the leftmost component of the fully-qualified host and domain name. For
example, if the hostname is ldap.exam ple.com , the instance name is ldap by default.
T he Admin Server directories are named the same as the Directory Server directories, only instead of
the instance as a directory name, the Admin Server directories are named adm in-serv. For any
25
Red Hat D irectory Server 8.2 Administration Guide
directory or folder named slapd-instance, substitute adm in-serv, such as /etc/dirsrv/slapdexam ple and /etc/dirsrv/adm in-serv.
T able 1.1. Red Hat Enterprise Linux 4 and 5 (x86)
File or Directory
Location
Log files
/var/log/dirsrv/slapd-instance
Configuration files
/etc/dirsrv/slapd-instance
Instance directory
/usr/lib/dirsrv/slapd-instance
Database files
/var/lib/dirsrv/slapd-instance
Runtime files
/var/lock/dirsrv/slapd-instance
/var/run/dirsrv/slapd-instance
Initscripts
/etc/rc.d/init.d/dirsrv and
/etc/sysconfig/dirsrv
/etc/rc.d/init.d/dirsrv-adm in and
/etc/sysconfig/dirsrv-adm in
T ools
/usr/bin/
/usr/sbin/
T able 1.2. Red Hat Enterprise Linux 4 and 5 (x86_64 )
File or Directory
Location
Log files
/var/log/dirsrv/slapd-instance
Configuration files
/etc/dirsrv/slapd-instance
Instance directory
/usr/lib64 /dirsrv/slapd-instance
Database files
/var/lib/dirsrv/slapd-instance
Runtime files
/var/lock/dirsrv/slapd-instance
/var/run/dirsrv/slapd-instance
Initscripts
/etc/rc.d/init.d/dirsrv and
/etc/sysconfig/dirsrv
/etc/rc.d/init.d/dirsrv-adm in and
/etc/sysconfig/dirsrv-adm in
T ools
/usr/bin/
/usr/sbin/
1.3. LDAP Tool Locations
Red Hat Directory Server uses Mozilla LDAP tools — such as ldapsearch, ldapm odify, and
ldapdelete — for command-line operations. T he MozLDAP tools are installed with Directory Server.
Platform
26
Directory Location
Chapter 1. Basic Red Hat D irectory Server Settings
Red Hat Enterprise Linux 4 and 5 i386
/usr/lib/mozldap
Red Hat Enterprise Linux 4 and 5 x86_64
/usr/lib64/mozldap
For all Red Hat Directory Server guides and documentation, the LDAP tools used in the examples, such
as ldapsearch and ldapm odify, are the Mozilla LDAP tools. For most Linux systems, OpenLDAP
tools are already installed in the /usr/bin/ directory. T hese OpenLDAP tools are not supported for
Directory Server operations. For the best results with the Directory Server, make sure the path to the
Mozilla LDAP tools comes first in the PAT H or use the full path and file name for every LDAP operation.
However, these OpenLDAP tools can be used for Directory Server operations with certain cautions:
T he output of the other tools may be different, so it may not look like the examples in the
documentation.
T he OpenLDAP tools require a -x argument to disable SASL so that it can be used for a simple bind,
meaning the -D and -w arguments or an anonymous bind.
T he OpenLDAP tools' arguments for using T LS/SSL and SASL are quite different than the Mozilla
LDAP arguments. See the OpenLDAP documentation for instructions on those arguments.
1.4. Starting and Stopping Servers
T he Directory Server is running when the setup-ds-adm in.pl script completes. Avoid stopping and
starting the server to prevent interrupting replication, searches, and other server operations.
If the Directory Server has SSL enabled, you cannot restart the server from the Console; you must
use the command-line. It is possible to restart without being prompted for a password; see
Section 14.2.3.3, “Creating a Password File for the Directory Server” for more information.
Rebooting the host system can automatically start the ns-slapd process. T he directory provides
startup or run command (rc) scripts. Use the chkconfig command to enable the Directory Server
and Admin Server to start on boot.
1.4.1. Starting and Stopping Directory Server from the Console
1. Start the Directory Server Console.
redhat-idm-console -a http://localhost:9830
2. In the T asks tab, click Start the Directory Server, Stop the Directory Server, or
Restart the Directory Server.
27
Red Hat D irectory Server 8.2 Administration Guide
When the Directory Server is successfully started or stopped from the Directory Server Console, the
server displays a message box stating that the server has either started or shut down.
1.4.2. Starting and Stopping Directory Server from the Command Line
T he most common way to start and stop the Directory Server service is using system tools on Red Hat
Enterprise Linux. For example, Linux uses the service tool:
service dirsrv {start|stop|restart} instance
Passing the instance name stops or starts only that instance; not giving any name starts or stops all
instances.
NOTE
T he service name for the Directory Server service on Red Hat Enterprise Linux is dirsrv.
T he start/stop scripts are in the /usr/sbin directory and are run similar to the service start/stop
command:
/usr/sbin/{start|stop|restart}-dirsrv instance
If the instance name is not given, then the all instances are started or stopped.
Alternatively each instance has its own start and stop scripts that apply only to that instance.
/etc/dirsrv/slapd-instance_name/{start|stop|restart}-slapd
1.4.3. Starting and Stopping Admin Server
T here are two ways to start, stop, or restart the Admin Server:
T here are scripts in the /usr/sbin directory.
/usr/sbin/{start|stop|restart}-ds-admin
28
Chapter 1. Basic Red Hat D irectory Server Settings
T he Admin Server service can also be stopped and started using system tools on Red Hat
Enterprise Linux. For example, on Red Hat Enterprise Linux 5 (64-bit), the command is service:
service dirsrv-admin {start|stop|restart}
NOTE
T he service name for the Admin Server process on Red Hat Enterprise Linux is dirsrvadm in.
1.5. Starting the Console
1.5.1. Starting the Directory Server Console
T here is a simple script to launch the Directory Server Console. T he script is in the standard /usr/bin
directory, so it can be run as follows:
redhat-idm-console
NOTE
Make sure that the correct Sun JDK or OpenJDK version is set in the PAT H before launching the
Console. Run the following to see if the Java program is in the PAT H and to get the version and
vendor information:
java -version
T he login screen prompts for the username, password, and Admin Server location. It is possible to pass
other information along with the Console command to supply the Admin Server URL, password, and
username. For example:
redhat-idm-console -a http://localhost:9830 -u "cn=Directory Manager" -w secret
29
Red Hat D irectory Server 8.2 Administration Guide
T able 1.3. redhat-idm-console Options
Option
Description
-a adminURL
Specifies a base URL for the instance of Admin Server to log into.
-f fileName
Writes errors and system messages to fileName.
-h
Prints out the help message for redhat-idm -console.
-s
Specifies the directory instance to access, either by specifying the DN of the server
instance entry (SIE) or the instance name, such as slapd-exam ple.
-u
Gives the user DN to use to log into the Console.
-w
Gives the password to use to log into the Console.
-w -
Reads the password from the standard output.
-x options
Specifies extra options. T here are three values for extraOptions:
nowinpos, which puts the Console window in the upper left corner of the screen
nologo, which keeps the splash screen from being displayed and only opens the
login dialog
javalaf, which uses the Java look and feel for the Console interface rather than the
platform-specific styles
T o use multiple options, separate them with a comma.
-y file
Reads the password from the specified input file.
1.5.2. Logging into Directory Server
After starting the Directory Server Console, a login screen opens, requiring the username and password
for the user logging in and the URL for the Admin Server instance being access. T he user logged in at
the Console is the user who is binding to Directory Server. T his determines the access permissions
granted and allowed operations while access the directory tree. T he user account used to log into the
Directory Server Console can make significant differences in the access; for example, the Directory
Manager has access to every user and configuration entry in Directory Server, while the adm in entry
created during installation has access to only configuration entries, not user entries. Regular user
accounts are more limited.
T o bind to, or log into, the Directory Server, supply a username and password at the login box.
30
Chapter 1. Basic Red Hat D irectory Server Settings
1.5.3. Changing Login Identity
At any time during a session, you can log in as a different user, without having to restart the Console. T o
change the login identity:
1. In the Directory Server Console, select the T asks tab.
2. Click Log on to the Directory Server as a New User.
3. A login dialog box appears.
Enter the full distinguished name of the entry with which to bind to the server. For example, to bind
as user Barbara Jensen, enter her full DN in the login box:
cn=Barbara Jensen,ou=People,dc=example,dc=com
1.5.4. Viewing the Current Console Bind DN
T o see the bind DN that is currently logged into the Directory Server Console, click the login icon in the
lower-left corner of the window. T he current bind DN appears next to the login icon.
Figure 1.1. Viewing the Bind DN
31
Red Hat D irectory Server 8.2 Administration Guide
1.6. Enabling LDAPI
Inter-process communication (IPC) is a way for separate processes on a Unix machine or a network to
communicate directly with each other. LDAPI allows LDAP connections to run over IPC connections,
meaning that LDAP operations can run over Unix sockets. T hese connections are much faster and more
secure than regular LDAP connections.
LDAPI is enabled through two configuration attributes:
nsslapd-ldapilisten to enable LDAPI for Directory Server
nsslapd-ldapifilepath to point to the Unix socket file
T o enable LDAPI:
1. Modify the nsslapd-ldapilisten to turn LDAPI on and add the socket file attribute.
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=config
changetype: modify
replace: nsslapd-ldapilisten
nsslapd-ldapilisten: on
add: nsslapd-ldapifilepath
nsslapd-ldapifilepath: /var/run/slapd-example.socket
2. Restart the server to apply the new configuration.
service dirsrv restart example
1.7. Changing Directory Server Port Numbers
T he standard and secure LDAP port numbers used by Directory Server can be changed through the
Directory Server Console or by changing the value of the nsslapd-port or nsslapd-secureport
attribute under the cn=config entry in the dse.ldif.
NOTE
Modifying the standard or secure port numbers for a Configuration Directory Server, which
maintains the o=NetscapeRoot subtree should be done through the Directory Server Console.
Changing the configuration directory or user directory port or secure port numbers has the following
repercussions:
T he Directory Server port number must also be updated in the Admin Server configuration.
If there are other Directory Server instances that point to the configuration or user directory, update
those servers to point to the new port number.
T o modify a Directory Server LDAP or LDAPS port for either a user or a configuration directory:
32
Chapter 1. Basic Red Hat D irectory Server Settings
1. In the Directory Server Console, select the Configuration tab, and then select the top entry in
the navigation tree in the left pane.
2. Select the Settings tab in the right pane.
3. Change the port numbers. T he port number for the server to use for non-SSL communications in
the Port field, with a default value of 389. T he port number for the server to use for SSL
communications is in the Encrypted Port field, with a default value of 636.
4. Click Save.
5. T he Console returns a warning, You are about to change the port number for the Configuration
Directory. This will affect all Administration Servers that use this directory and you'll need to update
them with the new port number. Are you sure you want to change the port number? Click Yes.
6. T hen a dialog appears, reading that the changes will not take effect until the server is restarted.
Click OK.
NOTE
Do not restart the Directory Server at this point. If you do, you will not be able to make the
necessary changes to the Admin Server through the Console.
7. Open the Admin Server Console.
8. In the Configuration tab, select the Configuration DS tab.
33
Red Hat D irectory Server 8.2 Administration Guide
9. In the LDAP Port field, type in the new LDAP port number for your Directory Server instance.
10. Check the Secure Connection box if this is a secure port.
NOTE
If you try to save these changes at this step, you will get a warning box that reads, Invalid
LDAP Host/LDAP Port, can not connect. Click OK, and ignore this warning.
11. In the T asks tab of the Directory Server Console, click Restart Directory Server. A dialog
to confirm that you want to restart the server. Click Yes.
12. Open the Configuration DS tab of the Admin Server Console and select Save.
A dialog will appear, reading The Directory Server setting has been modified. You must shutdown
and restart your Admin Server and all the servers in the Server Group for the changes to take
effect. Click OK.
13. In the T asks tab of the Admin Server Console, click Restart Adm in Server. A dialog opens
reading that the Admin Server has been successfully restarted. Click Close.
NOTE
You must close and reopen the Console before you can do anything else in the Console.
Refresh may not update the Console, and, if you try to do anything, you will get a warning
that reads Unable to contact LDAP server.
1.8. Creating a New Directory Server Instance
Additional instances can be created through the Directory Server Console or using the setup-ds.pl
script. For information on using the setup-ds.pl script, see the Directory Server Installation Guide. T o
create an instance using the Directory Server Console:
1. In the Red Hat Console window, select Server Group in the navigation tree, and then right-click.
2. From the pop-up menu, select Create Instance and then Directory Server.
34
Chapter 1. Basic Red Hat D irectory Server Settings
3. Fill in the instance information.
A unique name for the server. T his name must only have alphanumeric characters, a dash (-),
or an underscore (_).
A port number for LDAP communications.
T he root suffix for the new Directory Server instance.
A DN for the Directory Manager. T his user has total access to every entry in the directory,
without normal usage constraints (such as search timeouts); this is described in Section 1.9,
“Setting the Directory Manager Information”.
T he password for the Directory Manager.
T he user ID as which to run the Directory Server daemon.
4. Click OK.
A status box appears to confirm that the operation was successful. T o dismiss it, click OK.
1.9. Setting the Directory Manager Information
T he Directory Manager is the privileged database administrator, comparable to the root user in UNIX.
Access control does not apply to the Directory Manager entry; likewise, limits on searches and other
operations do not apply. T he Directory Manager entry is created during installation; the default DN is
cn=Directory Manager. T he password for this user is defined in the nsslapd-rootdn attribute.
T o change the Directory Manager DN and password and the encryption scheme used for this password:
1. Log in to the Directory Server Console as Directory Manager. For information on changing the bind
DN, see Section 1.5.3, “Changing Login Identity”.
2. In the Directory Server Console, select the Configuration tab, and then select the top entry in
35
Red Hat D irectory Server 8.2 Administration Guide
the navigation tree in the left pane.
3. Select the Manager tab in the right pane.
4. Change the Directory Manager settings:
Change the distinguished name for the Directory Manager in the Root DN field. T he default
value is cn=Directory Manager.
Set the storage scheme for the server to use to store the password for Directory Manager in
the Manager Password Encryption pull-down menu.
Enter a new password, and confirm it.
1.10. Using Directory Server Plug-ins
Directory Server has a number of default plug-ins which configure core Directory Server functions, such
as replication, classes of service, and even attribute syntaxes. Core plug-ins are enabled and completely
configured by default.
Other default plug-ins extend the functionality of the Directory Server by providing consistent, but userdefined, behaviors, as with DNA, attribute uniqueness, and attribute linking. T hese plug-ins are available,
but not enabled or configured by default.
Using plug-ins also allows the Directory Server to be easily extended, so customers can write and
deploy their own server plug-ins to perform whatever directory operations they need for their specific
deployment.
T he details of configuring and deploying plug-ins are covered in other guides (primarily the Plug-in
Programmer's Guide and to some extent in the plug-in attribute reference in the Configuration and
Command-Line Tool Reference). T his section covers common adminsitrative tasks for all plug-ins.
1.10.1. Enabling Plug-ins
T o enable and disable plug-ins over LDAP using the Directory Server Console:
1. In the Directory Server Console, select the Configuration tab.
2. Double-click the Plugins folder in the navigation tree.
3. Select the plug-in from the Plugins list.
4. T o disable the plug-in, clear the Enabled checkbox. T o enable the plug-in, check this checkbox.
36
Chapter 1. Basic Red Hat D irectory Server Settings
5. Click Save.
6. Restart the Directory Server.
service dirsrv restart instance_name
T o disable or enable a plug-in through the command line, use the ldapm odify utility to edit the value
of the nsslapd-pluginEnabled attribute. For example:
[1]
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=ACL Plugin,cn=plugins,cn=config
changetype: modify
replace: nsslapd-pluginEnabled
nsslapd-pluginEnabled: on
1.10.2. Setting the Plug-in Precedence
T he plug-in precedence is the priority it has in the execution order of plug-ins. For pre- and postoperation plug-ins, this allows one plug-in to be executed and complete before the next plug-in is
initiated, which lets the second plug-in take advantage of the first plug-in's results.
Plug-in precedence is configured in the nsslapd-pluginPrecedence attribute on the plug-in's
configuration entry. T his attribute has a value of 1 (highest priority) to 99 (lowest priority). If the attribute
is not set, it has a default value of 50.
T he nsslapd-pluginPrecedence attribute is set using the ldapm odify command. For example:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=My Example Plugin,cn=plugins,cn=config
changetype: modify
replace: nsslapd-pluginPrecedence
nsslapd-pluginPrecedence: 1
37
Red Hat D irectory Server 8.2 Administration Guide
IMPORTANT
Don't set the plug-in precedence for the default Directory Server plug-ins unless told to do so by
Red Hat support. T he plug-in precedence attribute is primarily to govern the behavior of custom
plug-ins, not to change the behavior of the core Directory Server plug-ins.
[1] The LDAP to o ls referenc ed in this g uid e are Mo z illa LDAP, ins talled with Direc to ry Server in the /usr/l i b 6 4 /m o zl d ap
d irec to ry o n Red Hat Enterp ris e Linux 5 (6 4-b it); d irec to ries fo r o ther p latfo rms are lis ted in Sec tio n 1.3, “ LDAP To o l Lo c atio ns ” .
Ho wever, Red Hat Enterp ris e Linux s ys tems als o inc lud e LDAP to o ls fro m O p enLDAP. It is p o s s ib le to us e the O p enLDAP
c o mmand s as s ho wn in the examp les , b ut yo u mus t us e the - x arg ument to d is ab le SASL and allo w s imp le authentic atio n.
38
Chapter 2. Configuring D irectory D atabases
Chapter 2. Configuring Directory Databases
T he directory is made up of databases, and the directory tree is distributed across the databases. T his
chapter describes how to create suffixes, the branch points for the directory tree, and how to create the
databases associated with each suffix. T his chapter also describes how to create database links to
reference databases on remote servers and how to use referrals to point clients to external sources of
directory data.
For a discussion of concepts about distributing directory data, see the Directory Server Deployment
Guide.
2.1. Creating and Maintaining Suffixes
Different pieces of the directory tree can be stored in different databases, and then these databases
can be distributed across multiple servers. T he directory tree contains branch points called nodes.
T hese nodes may be associated with databases. A suffix is a node of the directory tree associated with
a particular database. For example, a simple directory tree might appear as illustrated in Figure 2.1, “A
Directory T ree with One Root Suffix”.
Figure 2.1. A Directory T ree with One Root Suffix
T he ou=people suffix and all the entries and nodes below it might be stored in one database, the
ou=groups suffix on another database, and the ou=contractors suffix on yet another database.
Section 2.1.1, “Creating Suffixes”
Section 2.1.2, “Maintaining Suffixes”
2.1.1. Creating Suffixes
Both root and sub suffixes can be created to organize the contents of the directory tree. A root suffix is
the parent of a sub suffix. It can be part of a larger tree designed for the Directory Server. A sub suffix is
a branch underneath a root suffix. T he data for root and sub suffixes are contained by databases.
A directory might contain more than one root suffix. For example, an ISP might host several websites,
one for exam ple.com and one for redhat.com . T he ISP would create two root suffixes, one
corresponding to the dc=exam ple,dc=com naming context and one corresponding to the
dc=redhat,dc=com naming context, as shown in Figure 2.2, “A Directory T ree with T wo Root
Suffixes”.
39
Red Hat D irectory Server 8.2 Administration Guide
Figure 2.2. A Directory T ree with T wo Root Suffixes
It is also possible to create root suffixes to exclude portions of the directory tree from search operations.
For example, Example Corporation wants to exclude their European office from a search on the general
Example Corporation directory. T o do this, they create two root suffixes. One root suffix corresponds to
the general Example Corporation directory tree, dc=exam ple,dc=com , and one root suffix
corresponds to the European branch of their directory tree, l=europe,dc=exam ple,dc=com . From a
client application's perspective, the directory tree looks as illustrated in Figure 2.3, “A Directory T ree with
a Root Suffix Off Limits to Search Operations”.
Figure 2.3. A Directory T ree with a Root Suffix Off Limits to Search Operations
Searches performed by client applications on the dc=exam ple,dc=com branch of Example
Corporation's directory will not return entries from the l=europe,dc=exam ple,dc=com branch of the
directory, as it is a separate root suffix.
If Example Corporation decides to include the entries in the European branch of their directory tree in
general searches, they make the European branch a sub suffix of the general branch. T o do this, they
create a root suffix for Example Corporation, dc=exam ple,dc=com , and then create a sub suffix
beneath it for their European directory entries, l=europe,dc=exam ple,dc=com . From a client
application's perspective, the directory tree appears as illustrated in Figure 2.4, “A Directory T ree with a
Sub Suffix”.
40
Chapter 2. Configuring D irectory D atabases
Figure 2.4 . A Directory T ree with a Sub Suffix
T his section describes creating root and sub suffixes for the directory using either the Directory Server
Console or the command line.
Section 2.1.1.1, “Creating a New Root Suffix Using the Console”
Section 2.1.1.2, “Creating a New Sub Suffix Using the Console”
Section 2.1.1.3, “Creating Root and Sub Suffixes from the Command Line”
2.1.1.1. Creating a New Root Suffix Using the Console
1. In the Directory Server Console, select the Configuration tab.
2. Right-click Data in the left navigation pane, and select New Root Suffix from the pop-up menu.
3. Enter a unique suffix in the New suffix field.
T he suffix must be named with dc naming conventions, such as dc=exam ple,dc=com .
41
Red Hat D irectory Server 8.2 Administration Guide
4. Select the Create associated database autom atically to create a database at the
same time as the new root suffix, and enter a unique name for the new database in the Database
nam e field, such as exam ple2. T he name can be a combination of alphanumeric characters,
dashes (-), and underscores (_). No other characters are allowed.
Deselect the checkbox to create a database for the new root suffix later. T his option specifies a
directory where the database will be created. T he new root suffix will be disabled until a database
is created.
T he new root suffix is listed under the Data folder.
2.1.1.2. Creating a New Sub Suffix Using the Console
1. In the Directory Server Console, select the Configuration tab.
2. Under the Data in the left navigation pane, select the suffix under which to add a new sub suffix.
Right-click the suffix, and select New Sub Suffix from the pop-up menu.
42
Chapter 2. Configuring D irectory D atabases
T he Create new sub suffix dialog box is displayed.
3. Enter a unique suffix name in the New suffix field. T he suffix must be named in line with dc
naming conventions, such as ou=groups.
T he root suffix is automatically added to the name. For example, it the sub suffix ou=groups is
created under the dc=exam ple,dc=com suffix, the Console automatically names it
ou=groups,dc=exam ple,dc=com .
4. Select the Create associated database autom atically checkbox to create a database
at the same time as the new sub suffix, and enter a unique name for the new database in the
Database nam e field, such as exam ple2. T he name can be a combination of alphanumeric
characters, dashes (-), and underscores (_). No other characters are allowed.
If the checkbox is not selected, than the database for the new sub suffix must be created later.
T he new sub suffix is disabled until a database is created.
T he suffix appears automatically under its root suffix in the Data tree in the left navigation pane.
43
Red Hat D irectory Server 8.2 Administration Guide
2.1.1.3. Creating Root and Sub Suffixes from the Command Line
Use the ldapm odify command-line utility to add new suffixes to the directory configuration file. T he
suffix configuration information is stored in the cn=m apping tree,cn=config entry.
NOTE
Avoid creating entries under the cn=config entry in the dse.ldif file. T he cn=config entry
in the simple, flat dse.ldif configuration file is not stored in the same highly scalable database
as regular entries. As a result, if many entries, particularly entries that are likely to be updated
frequently, are stored under cn=config, performance will suffer.
1. Add a new root suffix to the configuration file using the ldapm odify utility.
[2]
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389
-h server.example.com
ldapm odify binds to the server and prepares it to add an entry to the configuration file.
2. Create the root suffix entry.
Example 2.1. Example Root Suffix Entry
dn: cn=dc=example\,dc=com,cn=mapping tree,cn=config
objectclass: top
objectclass: extensibleObject
objectclass: nsMappingTree
nsslapd-state: backend
nsslapd-backend: UserData
cn: dc=example,dc=com
3. Create a sub suffix for groups under this root suffix using ldapm odify to add the sub suffix
entry:
44
Chapter 2. Configuring D irectory D atabases
dn: cn=ou=groups\,dc=example\,dc=com,cn=mapping tree,cn=config
objectclass: top
objectclass: extensibleObject
objectclass: nsMappingTree
nsslapd-state: backend
nsslapd-backend: GroupData
nsslapd-parent-suffix: dc=example,dc=com
cn: ou=groups,dc=example,dc=com
NOTE
T o maintain suffixes using the Directory Server Console, respect the same spacing used to name
the root and sub suffixes in the command line. For example, if a root suffix is named ou=groups
,dc=exam ple,dc=com , with two spaces after groups, any sub suffixes created under this root
will need to specify two spaces after ou=groups, as well.
T he following table describes the attributes used to configure a suffix entry:
45
Red Hat D irectory Server 8.2 Administration Guide
T able 2.1. Suffix Attributes
Attribute Name
Value
dn
Defines the DN for the suffix. T he DN is
contained in quotes. T he value entered takes the
form
cn=dc=exam ple\,dc=com ,cn=m apping
tree,cn=config. T his attribute is required.
cn
Defines the relative DN (RDN) of the entry. T his
attribute is required.
objectclass
T ells the server that the entry is root or sub suffix
entry. It always takes the value
nsMappingT ree. T his attribute is required.
nsslapd-state
Determines how the suffix handles operations.
T his attribute takes the following values:
backend: T he backend (database) is used to
process all operations.
disabled: T he database is not available for
processing operations. T he server returns a
No such search object error in response to
requests made by client applications.
referral: A referral is returned for requests
made to this suffix.
referral on update: T he database is
used for all operations except update
requests, which receive a referral.
T he default value is disabled.
nsslapd-referral
Defines the LDAP URL of the referral to be
returned by the suffix. T his attribute can be multivalued, with one referral per value. T his attribute
is required when the value of the nsslapd-state
attribute is referral or referral on
update.
nsslapd-backend
Gives the name of the database or database link
used to process requests. T his attribute can be
multi-valued, with one database or database link
per value. See Section 2.3, “Creating and
Maintaining Database Links” for more information
about database links. T his attribute is required
when the value of the nsslapd-state attribute is
set to backend or referral on update.
nsslapd-distribution-plugin
Specifies the shared library to be used with the
custom distribution function. T his attribute is
required only when more than one database is
specified in the nsslapd-backend attribute. See
Section 2.2, “Creating and Maintaining
Databases” for more information about the
custom distribution function.
nsslapd-distribution-funct
Specifies the name of the custom distribution
46
Chapter 2. Configuring D irectory D atabases
function. T his attribute is required only when
more than one database is specified in the
nsslapd-backend attribute. See Section 2.2,
“Creating and Maintaining Databases” for more
information about the custom distribution function.
nsslapd-parent-suffix
Provides the DN of the parent entry for a sub
suffix. By default, this attribute is not present,
which means that the suffix is regarded as a root
suffix. For example, to create a sub suffix names
o=sales,dc=exam ple,dc=com under the root
suffix dc=exam ple,dc=com , add nsslapdparent-suffix: dc=exam ple,dc=com to
the sub suffix.
2.1.2. Maintaining Suffixes
T his section describes the following procedures:
Section 2.1.2.1, “Disabling a Suffix”
Section 2.1.2.2, “Deleting a Suffix”
2.1.2.1. Disabling a Suffix
Sometimes, a database may need taken down for maintenance, but the data the database contains are
not replicated. Rather than returning a referral, disable the suffix responsible for the database.
Once a suffix is disabled, the contents of the database related to the suffix are invisible to client
applications when they perform LDAP operations such as search, add, and modify.
T o disable a suffix:
1. In the Directory Server Console, select the Configuration tab.
2. Under Data in the left navigation pane, click the suffix to disable.
3. Click the Suffix Setting tab, and deselect the Enable this suffix checkbox.
47
Red Hat D irectory Server 8.2 Administration Guide
2.1.2.2. Deleting a Suffix
WARNING
Deleting a suffix also deletes all database entries and replication information associated with that
suffix.
1. In the Directory Server Console, select the Configuration tab.
2. Under Data in the left navigation pane, select the suffix to delete.
3. Right-click the suffix, and select Delete from the menu.
4. Select either Delete this suffix and all of its sub suffixes or Delete this
suffix only.
2.2. Creating and Maintaining Databases
After creating suffixes to organizing the directory data, create databases to contain that directory data.
Databases are used to store directory data.
2.2.1. Creating Databases
T he directory tree can be distributed over multiple Directory Server databases. T here are two ways to
distribute data across multiple databases:
One database per suffix. T he data for each suffix is contained in a separate database.
48
Chapter 2. Configuring D irectory D atabases
T hree databases are added to store the data contained in separate suffixes.
T his division of the tree corresponds to three databases.
Database one contains the data for ou=people plus the data for dc=exam ple,dc=com , so that
clients can conduct searches based at dc=exam ple,dc=com . Database two contains the data for
ou=groups, and database three contains the data for ou=contractors.
Multiple databases for one suffix.
Suppose the number of entries in the ou=people branch of the directory tree is so large that two
databases are needed to store them. In this case, the data contained by ou=people could be
distributed across two databases.
49
Red Hat D irectory Server 8.2 Administration Guide
DB1 contains people with names from A-K, and DB2 contains people with names from L-Z . DB3
contains the ou=groups data, and DB4 contains the ou=contractors data.
Custom distribution plug-in distributes data from a single suffix across multiple databases. Contact
Red Hat Professional Services for information on how to create distribution logic for Directory Server.
2.2.1.1. Creating a New Database for an Existing Suffix Using the Console
1. In the Directory Server Console, select the Configuration tab.
2. In the left pane, expand Data, then click the suffix to which to add the new database.
3. Right-click the suffix, and select New Database from the pop-up menu.
4. Enter a unique name for the database, such as exam ple2. T he database name can be a
combination of alphanumeric characters, dashes (-), and underscores (_).
50
Chapter 2. Configuring D irectory D atabases
T he Create database in field is automatically filled with the default database directory
(/var/lib/dirsrv/slapd-instance_name/db) and the name of the new database. It is also
possible to enter or browse for a different directory location.
2.2.1.2. Creating a New Database for a Single Suffix from the Command Line
Use the ldapm odify command-line utility to add a new database to the directory configuration file. T he
database configuration information is stored in the cn=ldbm database,cn=plugins,cn=config
entry.
For example, add a new database to the server exam ple1:
1. Run ldapm odify:[2]
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389
-h server.example.com
T he ldapm odify utility binds to the server and prepares it to add an entry to the configuration
file.
2. Create the entry for the new database.
dn: cn=UserData,cn=ldbm database,cn=plugins,cn=config
objectclass: extensibleObject
objectclass: nsBackendInstance
nsslapd-suffix: ou=people,dc=example,dc=com
T he entry added corresponds to a database named UserData that contains the data for the root
or sub suffix ou=people,dc=exam ple,dc=com .
3. Create a root or sub suffix, as described in Section 2.1.1.3, “Creating Root and Sub Suffixes from
the Command Line”. T he database name, given in the DN attribute, must correspond with the
value in the nsslapd-backend attribute of the suffix entry.
2.2.1.3. Adding Multiple Databases for a Single Suffix
A single suffix can be distributed across multiple databases. However, to distribute the suffix, a custom
distribution function has to be created to extend the directory. For more information on creating a custom
distribution function, contact Red Hat Professional Services.
51
Red Hat D irectory Server 8.2 Administration Guide
NOTE
Once entries have been distributed, they cannot be redistributed. T he following restrictions apply:
T he distribution function cannot be changed once entry distribution has been deployed.
T he LDAP m odrdn operation cannot be used to rename entries if that would cause them to
be distributed into a different database.
Distributed local databases cannot be replicated.
T he ldapm odify operation cannot be used to change entries if that would cause them to be
distributed into a different database.
Violating these restrictions prevents Directory Server from correctly locating and returning entries.
After creating a custom distribution logic plug-in, add it to the directory.
T he distribution logic is a function declared in a suffix. T his function is called for every operation
reaching this suffix, including subtree search operations that start above the suffix. A distribution
function can be inserted into a suffix using both the Console and the command line.
2.2.1.3.1. Adding the Custom Distribution Function to a Suffix Using the Directory Server
Console
1. In the Directory Server Console, select the Configuration tab.
2. Expand Data in the left navigation pane. Select the suffix to which to apply the distribution
function.
3. Select the Databases tab in the right window.
4. T he databases associated with the suffix are already listed in the Databases tab. Click Add to
associate additional databases with the suffix.
5. Enter the path to the distribution library.
6. Enter the name of the distribution function in the Function nam e field.
2.2.1.3.2. Adding the Custom Distribution Function to a Suffix Using the Command Line
1. Run ldapm odify.[2]
52
Chapter 2. Configuring D irectory D atabases
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
2. Add the following attributes to the suffix entry itself, supplying the information about the custom
distribution logic:
changetype: modify
add: nsslapd-backend
nsslapd-backend: Database1
add: nsslapd-backend
nsslapd-backend: Database2
add: nsslapd-backend
nsslapd-backend: Database3
add: nsslapd-distribution-plugin
nsslapd-distribution-plugin: /full/name/of/a/shared/library
add: nsslapd-distribution-funct
nsslapd-distribution-funct: distribution-function-name
T he nsslapd-backend attribute specifies all of the databases associated with this suffix. T he
nsslapd-distribution-plugin attribute specifies the name of the library that the plug-in uses.
T he nsslapd-distribution-funct attribute provides the name of the distribution function itself.
For more information about using the ldapm odify command-line utility, see Section 3.2.4, “Adding and
Modifying Entries Using ldapmodify”.
2.2.2. Maintaining Directory Databases
Section 2.2.2.1, “Placing a Database in Read-Only Mode”
Section 2.2.2.2, “Deleting a Database”
Section 2.2.2.3, “Configuring T ransaction Logs for Frequent Database Updates”
2.2.2.1. Placing a Database in Read-Only Mode
When a database is in read-only mode, you cannot create, modify, or delete any entries. One of the
situations when read-only mode is useful is for manually initializing a consumer or before backing up or
exporting data from the Directory Server. Read-only mode ensures a faithful image of the state of these
databases at a given time.
T he Directory Server Console and the command-line utilities do not automatically put the directory in
read-only mode before export or backup operations because this would make your directory unavailable
for updates. However, with multi-master replication, this might not be a problem.
Section 2.2.2.1.1, “Making a Database Read-Only Using the Console”
Section 2.2.2.1.2, “Making a Database Read-Only from the Command Line”
Section 2.2.2.1.3, “Placing the Entire Directory Server in Read-Only Mode”
2.2.2.1.1. Making a Database Read-Only Using the Console
1. In the Directory Server Console, select the Configuration tab.
2. Expand Data in the left pane. Expand the suffix containing the database to put in read-only mode.
53
Red Hat D irectory Server 8.2 Administration Guide
3. Select the database to put into read-only mode.
4. Select the Database Settings tab in the right pane.
5. Select the database is read-only checkbox.
T he change takes effect immediately.
Before importing or restoring the database, ensure that the databases affected by the operation are not
in read-only mode.
T o disable read-only mode, open the database up in the Directory Server Console again and uncheck
the database is read-only checkbox.
2.2.2.1.2. Making a Database Read-Only from the Command Line
T o manually place a database into read-only mode:
1. Run ldapm odify.[2]
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
2. Change the read-only attribute to on
dn: cn=database_name,cn=ldbm database,cn=plugins,cn=config
changetype: modify
replace: nsslapd-readonly
nsslapd-readonly: on
NOTE
By default, the name of the database created at installation time is userRoot.
2.2.2.1.3. Placing the Entire Directory Server in Read-Only Mode
If the Directory Server maintains more than one database and all databases need to be placed in readonly mode, this can be done in a single operation.
54
Chapter 2. Configuring D irectory D atabases
WARNING
T his operation also makes the Directory Server configuration read-only; therefore, you cannot
update the server configuration, enable or disable plug-ins, or even restart the Directory Server
while it is in read-only mode. Once read-only mode is enabled, it cannot cannot be undone from
the Console; you must modify the configuration files.
NOTE
If Directory Server contains replicas, do not use read-only mode because it will disable replication.
T o put the Directory Server in read-only mode:
1. In the Directory Server Console, select the Configuration tab, and then select the top entry in
the navigation tree in the left pane.
2. Select the Settings tab in the right pane.
3. Select the Make Entire Server Read-Only checkbox.
4. Click Save, and then restart the server.
2.2.2.2. Deleting a Database
Deleting a database deletes the configuration information and entries for that database only, not the
physical database itself.
1. In the Directory Server Console, select the Configuration tab.
2. Expand the Data folder, and then select the suffix.
3. Select the database to delete.
4. Right-click the database and select Delete from the pop-up menu.
55
Red Hat D irectory Server 8.2 Administration Guide
5. Confirm that the database should be deleted in the Delete Database dialog box.
2.2.2.3. Configuring T ransaction Logs for Frequent Database Updates
When the server is going to be asked to perform frequent database updates (LDAP adds, modifies,
replication), the database transaction log files should be configured to be on a different disk than the
primary database files.
Storing the transaction log files on a separate physical disk improves performance because the disk
heads do not thrash moving between the log files and the data files.
1. Stop the Directory Server instance.
service dirsrv stop example
2. Create the new directory, if necessary, where the transaction logs will be located.
mkdir /home/exampledb-txnlogs
3. Set the appropriate file permissions on the directory so that the Directory Server user can access
it; the default Directory Server user and group are nobody:nobody.
chown nobody:nobody /home/exampledb-txnlogs
4. Open the Directory Server instance's configuration directory.
cd /etc/dirsrv/slapd-instance_name
5. Edit the dse.ldif file, and change the nsslapd-db-logdirectory directive for the new log
file path:
nsslapd-db-logdirectory: /home/exampledb-txnlogs
T his attribute goes on the same entry that has the nsslapd-dbcachesize attribute.
6. Open the database directory.
56
Chapter 2. Configuring D irectory D atabases
cd /var/lib/dirsrv/slapd-instance_name/db
7. Remove all of the __db.* files.
8. Move the log.* files to the new location.
9. Start the Directory Server instance again.
service dirsrv start example
2.2.3. Configuring Attribute Encryption
T he Directory Server offers a number of mechanisms to secure access to sensitive data, such as
access control rules to prevent unauthorized users from reading certain entries or attributes within
entries and SSL to protect data from eavesdropping and tampering on untrusted networks. However, if a
copy of the server's database files should fall into the hands of an unauthorized person, they could
potentially extract sensitive information from those files. Because information in a database is stored in
plain text, some sensitive information, such as government identification numbers or passwords, may not
be protected enough by standard access control measures.
For highly sensitive information, this potential for information loss could present a significant security
risk. In order to remove that security risk, Directory Server allows portions of its database to be
encrypted. Once encrypted, the data are safe even in the event that an attacker has a copy of the
server's database files.
Database encryption allows attributes to be encrypted in the database. Both encryption and the
encryption cipher are configurable per attribute per backend. When configured, every instance of a
particular attribute, even index data, is encrypted for every entry stored in that database.
NOTE
T o enable attribute encryption on an attribute with existing stored data, export the database to
LDIF first, then make the configuration change, then re-import the data to the database. T he
server does not enforce consistency between encryption configuration and stored data; therefore,
pay careful attention that all existing data are exported before enabling or disabling encryption.
Indexed attributes may be encrypted, and attribute encryption is fully compatible with indexing. T he
contents of the index files that are normally derived from attribute values are also encrypted to prevent
an attacker from recovering part or all of the encrypted data from an analysis of the indexes.
Since the server pre-encrypts all index keys before looking up an index for an encrypted attribute, there
is some effect on server performance for searches that make use of an encrypted index, but the effect is
not serious enough that it is no longer worthwhile to use an index.
2.2.3.1. Encryption Keys
In order to use attribute encryption, the server must be configured for SSL and have SSL enabled
because attribute encryption uses the server's SSL encryption key and the same PIN input methods as
SSL. T he PIN must either be entered manually upon server startup or a PIN file must be used.
Randomly generated symmetric cipher keys are used to encrypt and decrypt attribute data. A separate
key is used for each configured cipher. T hese keys are wrapped using the public key from the server's
SSL certificate, and the resulting wrapped key is stored within the server's configuration files. T he
effective strength of the attribute encryption is never higher than the strength of the server's SSL key
used for wrapping. Without access to the server's private key, it is not possible to recover the symmetric
57
Red Hat D irectory Server 8.2 Administration Guide
keys from the wrapped copies.
WARNING
T here is no mechanism for recovering a lost key. T herefore, it is especially important to back up
the server's certificate database safely. If the server's certificate were lost, it would not be
possible to decrypt any encrypted data stored in its database.
WARNING
If the SSL certificate is expiring and needs to be renewed, export the encrypted backend instance
before the renewal. Update the certificate, then re-import the exported LDIF file.
2.2.3.2. Encryption Ciphers
T he encryption cipher is configurable on a per-attribute basis and must be selected by the administrator
at the time encryption is enabled for an attribute. Configuration can be done through the Console or
through the command line.
T he following ciphers are supported:
Advanced Encryption Standard (AES)
T riple Data Encryption Standard (3DES)
All ciphers are used in Cipher Block Chaining mode.
Once the encryption cipher is set, it should not be changed without exporting and re-importing the data.
2.2.3.3. Configuring Attribute Encryption from the Console
1. In the Configuration tab, select the Data node.
2. Expand the suffix, and select the database to edit.
3. Select the Attribute Encryption tab.
4. Click the Add Attribute button to open the list of attributes. Select the attribute to encrypt.
58
Chapter 2. Configuring D irectory D atabases
NOTE
For existing attribute values to be encrypted, the information must be exported from the
database, then re-imported. See Section 2.2.3.5, “Exporting and Importing an Encrypted
Database”.
5. Select which encryption cipher to use.
NOTE
T he encryption cipher to use is set separately for each attribute, so attribute encryption is applied
to each attribute one at a time.
T o remove encryption from attributes, select them from the list of encrypted attributes in the Attribute
Encryption table, click the Delete button, and then click Save to apply the changes. Any deleted
attributes have to be manually re-added after saving.
59
Red Hat D irectory Server 8.2 Administration Guide
2.2.3.4 . Configuring Attribute Encryption Using the Command Line
1. Run the ldapm odify command[2]:
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389
-h server.example.com
2. Add an encryption entry for the attribute being encrypted. For example, this entry encrypts the
telephoneNumber attribute with the AES cipher:
dn: cn=telephoneNumber,cn=encrypted attributes,cn=Database1,cn=ldbm
database,cn=plugins,cn=config
objectclass: top
objectclass: nsAttributeEncryption
cn: telephoneNumber
nsEncryptionAlgorithm: AES
3. For existing attributes in entries to be encrypted, the information must be exported, then reimported. See Section 2.2.3.5, “Exporting and Importing an Encrypted Database”.
For more information on attribute encryption configuration schema, refer to "Database Attributes under
cn=attributeName,cn=encrypted attributes,cn=database_name,cn=ldbm database,cn=plugins,cn=config"
in the Directory Server Configuration and Command-Line Tool Reference.
2.2.3.5. Exporting and Importing an Encrypted Database
Exporting and importing encrypted databases is similar to exporting and importing regular databases.
However, the encrypted information must be decrypted when it is exported to LDIF, then re-encrypted
when it is imported to the database. Using the -E option when running the db2ldif and ldif2db
scripts will decrypt the data on export and re-encrypt it on import.
1. Export the data using the db2ldif script, as follows:
db2ldif -n Database1 -E -a /path/to/output.ldif -s "dc=example,dc=com" -s
"o=userRoot"
See Section 4.2.3, “Exporting to LDIF from the Command Line” for more information.
2. Make any configuration changes.
3. Re-import the data using the ldif2db script, as follows:
ldif2db -n Database1 -E -i /path/to/output.ldif
See Section 4.1.4, “Importing from the Command Line” for more information.
60
Chapter 2. Configuring D irectory D atabases
NOTE
When enabling encryption for data that is already present in the database, several additional
security concerns arise:
It is possible for old, unencrypted data to persist in the server's database page pool backing
file, even after a successful re-import with encryption. T o remove this data, stop the server
and delete the db/guardian file, then re-start the server. T his will force recovery, a sideeffect of which is deleting the backing file. However, it is possible that the data from the
deleted file could still be recovered from the hard drive unless steps are taken to overwrite the
disk blocks that it occupied.
After enabling encryption and importing data, be sure to delete the LDIF file because it
contains plain text values for the now-encrypted data. Ensure that the disk blocks that it
occupied are overwritten.
T he unencrypted data previously stored in the server's database may persist on disk after a
successful re-import with encryption. T his is because the old database files are deleted as
part of the import process. Ensure that the disk blocks that those files occupied are
overwritten.
Data stored in the server's replication log database is never encrypted; therefore, care should
be taken to protect those files if replication is used.
T he server does not attempt to protect unencrypted data stored in memory. T his data may be
copied into a system page file by the operating system. For this reason, ensure that any page
or swap files are adequately protected.
2.2.3.6. Updating Attribute Encryption Keys for New SSL/T LS Certificates
When SSL is first configured, there is no problem with attribute encryption. However, if the SSL certificate
is changed, then attribute encryption fails, with messages like these:
Apr 4 13:00:35 smtp1 logger: [04/Apr/2010:13:00:34 -0700]
attrcrypt_unwrap_key: failed to unwrap key for cipher AES
Apr 4 13:00:35 smtp1 logger: [04/Apr/2010:13:00:34 -0700]
key for cipher AES in attrcrypt_cipher_init
Apr 4 13:00:35 smtp1 logger: [04/Apr/2010:13:00:34 -0700]
cipher AES in attrcrypt_init
Apr 4 13:00:35 smtp1 logger: [04/Apr/2010:13:00:34 -0700]
attrcrypt_unwrap_key: failed to unwrap key for cipher AES
Apr 4 13:00:35 smtp1 logger: [04/Apr/2010:13:00:34 -0700]
key for cipher AES in attrcrypt_cipher_init
Apr 4 13:00:35 smtp1 logger: [04/Apr/2010:13:00:34 -0700]
cipher AES in attrcrypt_init
- Failed to retrieve
- Failed to initialize
- Failed to retrieve
- Failed to initialize
T his is because the previously-generated keys do not work with the new server certificate. T o correct
these errors, force the server to generate new keys for attribute encryption:
1. Stop the server.
service dirsrv stop
2. Open the dse.ldif file.
vim /etc/dirsrv/dse.ldif
61
Red Hat D irectory Server 8.2 Administration Guide
3. T here are special encryption key entries for the encryption ciphers used for attribute encryption
under the database configuration. For example:
dn: cn=AES,cn=encrypted attribute keys,cn=userRoot,cn=ldbm
database,cn=plugins,cn=config
objectClass: top
objectClass: extensibleObject
cn: AES
nssymmetrickey:: mSLm/RlCLvPZrSdARHPowedF9zKx+kjVTww5ARE4w0lbl2YlYvrI3bNg=
Delete these entries.
4. Start the server again.
service dirsrv start
2.3. Creating and Maintaining Database Links
Chaining means that a server contacts other servers on behalf of a client application and then returns
the combined results. Chaining is implemented through a database link, which points to data stored
remotely. When a client application requests data from a database link, the database link retrieves the
data from the remote database and returns it to the client.
Section 2.3.1, “Creating a New Database Link”
Section 2.3.2, “Configuring the Chaining Policy”
Section 2.3.3, “Maintaining Database Links”
Section 2.3.4, “Configuring Database Link Defaults”
Section 2.3.5, “Deleting Database Links”
Section 2.3.6, “Database Links and Access Control Evaluation”
For more general information about chaining, refer to the chapter "Designing the Directory T opology," in
the Directory Server Deployment Guide. Section 15.7, “Monitoring Database Link Activity” covers how to
monitor database link activity.
2.3.1. Creating a New Database Link
T he basic database link configuration requires four piece of information:
Suffix information. A suffix is created in the directory tree that is managed by the database link, not a
regular database. T his suffix corresponds to the suffix on the remote server that contains the data.
Bind credentials. When the database link binds to a remote server, it impersonates a user, and this
specifies the DN and the credentials for each database link to use to bind with remote servers.
LDAP URL. T his supplies the LDAP URL of the remote server to which the database link connects.
List of failover servers. T his supplies a list of alternative servers for the database link to contact in
the event of a failure. T his configuration item is optional.
NOTE
If secure binds are required for simple password authentication (Section 13.5.1, “Requiring
Secure Binds”), then any chaining operations will fail unless they occur over a secure connection.
Using a secure connection (SSL/T LS and Start T LS connections or SASL authentication) is
recommended, anyway.
62
Chapter 2. Configuring D irectory D atabases
2.3.1.1. Creating a New Database Link Using the Console
1. In the Directory Server Console, select the Configuration tab.
2. Create a new suffix as described in Section 2.1.1, “Creating Suffixes”.
Deselect the Create associated database autom atically checkbox. It is simpler to
configure a database link on a suffix without a database associated with it because having both a
database and database link requires custom distribution functions to distribute directory data.
3. In the left pane, right-click the new suffix, and select New Database Link from the pop-up menu.
4. Fill in the database link name. T he name can be a combination of alphanumeric characters,
dashes (-), and underscores (_). No other characters, like spaces, are allowed.
5. Set the radio button for the appropriate method for authentication.
63
Red Hat D irectory Server 8.2 Administration Guide
T here are four authentication methods:
Simple means that the server connects over the standard port with no encryption. T he only
required information is the bind DN and password for the user as whom the server connects to
the remote server.
Server TLS/SSL Certificate uses the local server's SSL certificate to authenticate to the remote
server. A certificate must be installed on the local server for certificate-based authentication,
and the remote server must have certificate mapping configured so that it can map the subject
DN in the local server's certificate to the corresponding user entry.
Configuring SSL and certificate mapping is described in Section 14.2, “Using T LS/SSL”.
NOTE
When the database link and remote server are configured to communicate using SSL,
this does not mean that the client application making the operation request must also
communicate using SSL. T he client can bind using a normal port.
SASL/DIGEST-MD5 requires only the bind DN and password to authenticate.
SASL/GSSAPI requires the local server to have a Kerberos keytab (as in Section 14.3.4.2,
“About the KDC Server and Keytabs”), and the remote server to have a SASL mapping to map
64
Chapter 2. Configuring D irectory D atabases
the local server's principal to the real user entry (as in Section 14.3.5.1, “Configuring SASL
Identity Mapping from the Console”).
6. In the Rem ote Server Inform ation section, select the connection type for the local server to
use to connect to the remote server. T here are three options:
Use LDAP. T his sets a standard, unencrypted connection.
Use TLS/SSL. T his uses a secure connection over the server's secure LDAPS port, such as
636. T his setting is required to use T LS/SSL.
When using SSL/T LS, make sure that the remote server's port number is set to its secure port.
Use Start TLS. T his uses Start T LS to establish a secure connection over the server's
standard port.
NOTE
If secure binds are required for simple password authentication (Section 13.5.1, “Requiring
Secure Binds”), then any chaining operations will fail unless they occur over a secure
connection. Using a secure connection (SSL/T LS and Start T LS connections or SASL
authentication) is recommended, anyway.
7. In the Rem ote Server Inform ation section, fill in the name and port number for the remote
server.
For any failover servers, fill in the hostname and port number, and click the Add button. A failover
server is a backup server, so that if the primary remote server fails, the database link contacts the
first server in the failover servers list and cycles through the list until a server is accessed.
T he new database link is listed under the suffix, in place of the database.
NOTE
T he Console provides a checklist of information that needs to be present on the remote server
for the database link to bind successfully. T o view this checklist, click the new database link, and
click the Authentication tab. T he checklist is in the Rem ote server checklist box.
2.3.1.2. Creating a Database Link from the Command Line
1. Use the ldapm odify command-line utility to create a new database link from the command line.
T he new instance must be located in the cn=chaining database,cn=plugins,cn=config
entry.
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389
-h server.example.com
2. Specify the configuration information for the database link:
65
Red Hat D irectory Server 8.2 Administration Guide
dn: cn=examplelink,cn=chaining database,cn=plugins,cn=config
objectclass: top
objectclass: extensibleObject
objectclass: nsBackendInstance
nsslapd-suffix: ou=people,dc=example,dc=com suffix being chained
nsfarmserverurl: ldap://people.example.com:389/ LDAP URL to remote server
nsmultiplexorbinddn: cn=proxy admin,cn=config bind DN
nsmultiplexorcredentials: secret bind password
cn: examplelink
NOTE
If secure binds are required for simple password authentication (Section 13.5.1, “Requiring
Secure Binds”), then any chaining operations will fail unless they occur over a secure connection.
Using a secure connection (SSL/T LS and Start T LS connections or SASL authentication) is
recommended, anyway.
Default configuration attributes are contained in the cn=default instance config,cn=chaining
database,cn=plugins,cn=config entry. T hese configuration attributes apply to all database links
at creation time. Changes to the default configuration only affect new database links. T he default
configuration attributes on existing database links cannot be changed.
Each database link contains its own specific configuration information, which is stored with the database
link entry itself, cn=database_link, cn=chaining database,cn=plugins,cn=config. For more
information about configuration attributes, refer to the Directory Server Configuration and Command-Line
Tool Reference.
Section 2.3.1.2.1, “Providing Suffix Information”
Section 2.3.1.2.2, “Providing Bind Credentials”
Section 2.3.1.2.3, “Providing an LDAP URL”
Section 2.3.1.2.4, “Providing a List of Failover Servers”
Section 2.3.1.2.5, “Using Different Bind Mechanisms”
Section 2.3.1.2.6, “Summary of Database Link Configuration Attributes”
Section 2.3.1.2.7, “Database Link Configuration Example”
2.3.1.2.1. Providing Suffix Information
Use the nsslapd-suffix attribute to define the suffix managed by the database link. For example, for
the database link to point to the people information for a remote site of the company, enter the following
suffix information:
nsslapd-suffix: l=Zanzibar,ou=people,dc=example,dc=com
T he suffix information is stored in the cn=database_link, cn=chaining
database,cn=plugins,cn=config entry.
66
Chapter 2. Configuring D irectory D atabases
NOTE
After creating the database link, any alterations to the nsslapd-nsslapd-suffix attribute are
applied only after the server containing the database link is restarted.
2.3.1.2.2. Providing Bind Credentials
For a request from a client application to be chained to a remote server, special bind credentials can be
supplied for the client application. T his gives the remote server the proxied authorization rights needed
to chain operations. Without bind credentials, the database link binds to the remote server as
anonym ous.
Providing bind credentials involves the following steps:
1. On the remote server:
Create an administrative user for the database link.
For information on adding entries, see Chapter 3, Creating Directory Entries.
Provide proxy access rights for the administrative user created in step 1 on the subtree
chained to by the database link.
For more information on configuring ACIs, see Chapter 12, Managing Access Control
2. On the server containing the database link, use ldapm odify to provide a user DN for the
database link in the nsMultiplexorBindDN attribute of the cn=database_link, cn=chaining
database,cn=plugins,cn=config entry.
WARNING
T he nsMultiplexorBindDN cannot be that of the Directory Manager.
Use ldapm odify to provide a user password for the database link in the
nsMultiplexorCredentials attribute of the cn=database_link, cn=chaining
database,cn=plugins,cn=config entry.
For example, a client application sends a request to Server A. Server A contains a database link that
chains the request to a database on Server B.
T he database link on Server A binds to Server B using a special user as defined in the
nsMultiplexorBindDN attribute and a user password as defined in the nsMultiplexorCredentials
attribute. In this example, Server A uses the following bind credentials:
67
Red Hat D irectory Server 8.2 Administration Guide
nsMultiplexorBindDN: cn=proxy admin,cn=config
nsMultiplexorCredentials: secret
Server B must contain a user entry corresponding to the nsMultiplexorBindDN, and set the proxy
authentication rights for this user. T o set the proxy authorization correctly, set the proxy ACI as any
other ACI.
WARNING
Carefully examine access controls when enabling chaining to avoid giving access to restricted
areas of the directory. For example, if a default proxy ACI is created on a branch, the users that
connect via the database link will be able to see all entries below the branch. T here may be
cases when not all of the subtrees should be viewed by a user. T o avoid a security hole, create
an additional ACI to restrict access to the subtree.
For more information on ACIs, see Chapter 12, Managing Access Control. For more information about the
proxy authentication control, refer to the LDAP C-SDK documentation at http://www.mozilla.org/directory.
NOTE
When a database link is used by a client application to create or modify entries, the attributes
creatorsName and modifiersName do not reflect the real creator or modifier of the entries.
T hese attributes contain the name of the administrative user granted proxied authorization rights
on the remote data server.
2.3.1.2.3. Providing an LDAP URL
On the server containing the database link, identify the remote server that the database link connects
with using an LDAP URL. Unlike the standard LDAP URL format, the URL of the remote server does not
specify a suffix. It takes the form ldap://hostname:port.
T he URL of the remote server using the nsFarmServerURL attribute is set in the cn=database_link,
cn=chaining database,cn=plugins,cn=config entry of the configuration file.
nsFarmServerURL: ldap://example.com:389/
68
Chapter 2. Configuring D irectory D atabases
NOTE
Do not forget to use the trailing slash (/) at the end of the URL.
For the database link to connect to the remote server using LDAP over SSL, the LDAP URL of the remote
server uses the protocol LDAPS instead of LDAP in the URL and points to the secure port of the server.
For example:
nsFarmServerURL: ldaps://africa.example.com:636/
NOTE
SSL has to be enabled on the local Directory Server and the remote Directory Server to be
chained over SSL. For more information on enabling SSL, see Section 14.2.1, “Enabling T LS/SSL:
Summary of Steps”.
When the database link and remote server are configured to communicate using SSL, this does
not mean that the client application making the operation request must also communicate using
SSL. T he client can bind using a normal port.
2.3.1.2.4 . Providing a List of Failover Servers
T here can be additional LDAP URLs for servers included to use in the case of failure. Add alternate
servers to the nsFarmServerURL attribute, separated by spaces.
nsFarmServerURL: ldap://example.com us.example.com:389 africa.example.com:1000/
In this sample LDAP URL, the database link first contacts the server exam ple.com on the standard
port to service an operation. If it does not respond, the database link then contacts the server
us.exam ple.com on port 389. If this server fails, it then contacts africa.exam ple.com on port
1000.
2.3.1.2.5. Using Different Bind Mechanisms
T he local server can connect to the remote server using several different connection types and
authentication mechanisms.
T here are three ways that the local server can connect to the remote server:
Over the standard LDAP port
Over a dedicated T LS/SSL port
Using Start T LS, which is a secure connection over a standard port
NOTE
If secure binds are required for simple password authentication (Section 13.5.1, “Requiring
Secure Binds”), then any chaining operations will fail unless they occur over a secure connection.
Using a secure connection (SSL/T LS and Start T LS connections or SASL authentication) is
recommended, anyway.
69
Red Hat D irectory Server 8.2 Administration Guide
Ultimately, there are two connection settings. T he T LS/SSL option signifies that both of the servers are
configured to run and accept connections over T LS/SSL, but there is no separate configuration attribute
for enforcing T LS/SSL.
T he connection type is identified in the nsUseStartTLS attribute. When this is on, then the server
initiates a Start T LS connect over the standard port. If this is off, then the server either uses the LDAP
port or the T LS/SSL port, depending on what is configured for the remote server in the
nsFarmServerURL attribute.
For example, to use Start T LS:
nsUseStartTLS: on
For example, to use a standard connection or T LS/SSL connection:
nsUseStartTLS: off
T here are four different methods which the local server can use to authenticate to the farm server.
empty. If there is no bind mechanism set, then the server performs simple authentication and
requires the nsMultiplexorBindDN and nsMultiplexorCredentials attributes to give the bind
information.
EXTERNAL. T his uses an SSL certificate to authenticate the farm server to the remote server. Either
the farm server URL must be set to the secure URL (ldaps) or the nsUseStartTLS attribute must
be set to on.
Additionally, the remote server must be configured to map the farm server's certificate to its bind
identity, as described in Section 14.2.10.2, “Mapping DNs to Certificates”.
DIGEST-MD5. T his uses SASL authentication with DIGEST -MD5 encryption. As with simple
authentication, this requires the nsMultiplexorBindDN and nsMultiplexorCredentials
attributes to give the bind information.
GSSAPI. T his uses Kerberos-based authentication over SASL.
T he farm server must be configured with a Kerberos keytab, and the remote server must have a
defined SASL mapping for the farm server's bind identity. Setting up Kerberos keytabs and SASL
mappings is described in Section 14.3, “Using SASL”.
NOTE
SASL connections can be established over standard connections or SSL/T LS connections.
For example:
nsBindMechanism: EXTERNAL
70
Chapter 2. Configuring D irectory D atabases
NOTE
If SASL is used, then the local server must also be configured to chain the SASL and password
policy components. Add the components for the database link configuration, as described in
Section 2.3.2, “Configuring the Chaining Policy”. For example:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=config,cn=chaining database,cn=plugins,cn=config
changetype: modify
add: nsActiveChainingComponents
nsActiveChainingComponents: cn=password policy,cn=components,cn=config
add: nsActiveChainingComponents
nsActiveChainingComponents: cn=sasl,cn=components,cn=config
^D
2.3.1.2.6. Summary of Database Link Configuration Attributes
T he following table lists the attributes available for configuring a database link. Some of these attributes
were discussed in the earlier sections. All instance attributes are defined in the cn=database_link,
cn=chaining database,cn=plugins,cn=config entry.
Values defined for a specific database link take precedence over the global attribute value.
71
Red Hat D irectory Server 8.2 Administration Guide
T able 2.2. Database Link Configuration Attributes
Attributes
nsT ransmittedControls
Value
[†]
Gives the OID of LDAP controls forwarded by the
database link to the remote data server.
nsslapd-suffix
T he suffix managed by the database link. Any
changes to this attribute after the entry has been
created take effect only after the server
containing the database link is restarted.
nsslapd-timelimit
Default search time limit for the database link,
given in seconds. T he default value is 3600
seconds.
nsslapd-sizelimit
Default size limit for the database link, given in
number of entries. T he default value is 2000
entries.
nsFarmServerURL
Gives the LDAP URL of the remote server (or
farm server) that contains the data. T his attribute
can contain optional servers for failover,
separated by spaces. If using cascading chaining,
this URL can point to another database link.
nsUseStartT LS
Sets whether to use Start T LS to establish a
secure connection over a standard port. T he
default is off, which is used for both simple
(standard) connections and T LS/SSL
connections.
nsBindMethod
Sets the authentication method to use to
authenticate (bind) to the remote server. T he
default, if this attribute is not given, is to
authenticate using a simple bind, requiring the
nsMultiplexorBindDN and
nsMultiplexorCredentials attributes for the
bind information.
null; this is a simple bind.
EXT ERNAL (certificate-based); certificate
mapping must be enabled for this.
DIGEST -MD5 (SASL); this, like a simple bind,
requires the bind DN and password.
GSSAPI (SASL); this requires the keytab to be
configured on the local server and SASL identity
mapping on the remote server.
nsMultiplexorBindDN
72
DN of the administrative entry used to
communicate with the remote server. T he term
multiplexor in the name of the attribute means
the server which contains the database link and
communicates with the remote server. T his bind
DN cannot be the Directory Manager. If this
attribute is not specified, the database link binds
Chapter 2. Configuring D irectory D atabases
as anonym ous.
nsMultiplexorCredentials
Password for the administrative user, given in
plain text. If no password is provided, it means
that users can bind as anonym ous. T he
password is encrypted in the configuration file.
nsCheckLocalACI
Reserved for advanced use only. Controls
whether ACIs are evaluated on the database link
as well as the remote data server. T akes the
values on or off. Changes to this attribute occur
only after the server has been restarted. T he
default value is off.
nsProxiedAuthorization
Reserved for advanced use only. Disables
proxied authorization. A value of off means
proxied authorization is disabled. T he default
value is on.
nsActiveChainingComponents [†]
Lists the components using chaining. A
component is any functional unit in the server.
T he value of this attribute in the database link
instance overrides the value in the global
configuration attribute. T o disable chaining on a
particular database instance, use the value none.
T he default policy is not to allow chaining. For
more information, see Section 2.3.2.1, “Chaining
Component Operations”.
nsReferralOnScopedSearch
Controls whether referrals are returned by
scoped searches. T his attribute is for optimizing
the directory because returning referrals in
response to scoped searches is more efficient.
T akes the values on or off. T he default value is
off.
nsHopLimit
Maximum number of times a request can be
forwarded from one database link to another. T he
default value is 10.
[†] Can b e b o th a g lo b al and ins tanc e attrib ute. This g lo b al c o nfig uratio n attrib ute is lo c ated in the cn = co n f i g ,cn = ch ai n i n g
d atab ase,cn = p l ug i n s,cn = co n f i g entry. The g lo b al attrib utes are d ynamic , meaning any c hang es mad e to them auto matic ally
take effec t o n all ins tanc es o f the d atab as e link within the d irec to ry.
2.3.1.2.7. Database Link Configuration Example
Suppose a server within the us.exam ple.com domain contains the subtree l=Walla
Walla,ou=people,dc=exam ple,dc=com on a database and that operation requests for the
l=Zanzibar,ou=people,dc=exam ple,dc=com subtree should be chained to a different server in
the africa.exam ple.com domain.
73
Red Hat D irectory Server 8.2 Administration Guide
1. Run ldapm odify[2] to add a database link to Server A:
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389
-h server.example.com
2. Specify the configuration information for the database link:
dn: cn=DBLink1,cn=chaining database,cn=plugins,cn=config
objectclass: top
objectclass: extensibleObject
objectclass: nsBackendInstance
nsslapd-suffix: c=africa,ou=people,dc=example,dc=com
nsfarmserverurl: ldap://africa.example.com:389/
nsmultiplexorbinddn: cn=proxy admin,cn=config
nsmultiplexorcredentials: secret
cn: DBLink1
dn: cn=c=africa\,ou=people\,dc=example\,dc=com,cn=mapping tree,cn=config
objectclass: top
objectclass: extensibleObject
objectclass: nsMappingTree
nsslapd-state: backend
nsslapd-backend: DBLink1
nsslapd-parent-suffix: ou=people,dc=example,dc=com
cn: c=africa\,ou=people\,dc=example\,dc=com
In the first entry, the nsslapd-suffix attribute contains the suffix on Server B to which to chain
from Server A. T he nsFarmServerURL attribute contains the LDAP URL of Server B.
T he second entry creates a new suffix, allowing the server to route requests made to the new
database link. T he cn attribute contains the same suffix specified in the nsslapd-suffix
attribute of the database link. T he nsslapd-backend attribute contains the name of the database
link. T he nsslapd-parent-suffix attribute specifies the parent of this new suffix,
74
Chapter 2. Configuring D irectory D atabases
ou=people,dc=exam ple,dc=com .
3. Create an administrative user on Server B, as follows:
dn: cn=proxy admin,cn=config
objectclass: person
objectclass: organizationalPerson
objectclass: inetOrgPerson
cn: proxy admin
sn: proxy admin
userPassword: secret
description: Entry for use by database links
WARNING
Do not use the Directory Manager user as the proxy administrative user on the remote
server. T his creates a security hole.
4. Add the following proxy authorization ACI to the
l=Zanzibar,ou=people,dc=exam ple,dc=com entry on Server B:
aci: (targetattr = "*")(version 3.0; acl "Proxied authorization
for database links"; allow (proxy) userdn = "ldap:///cn=proxy
admin,cn=config";)
T his ACI gives the proxy admin user read-only access to the data contained on the remote server
within the l=Zanzibar,ou=people,dc=exam ple,dc=com subtree only.
NOTE
When a user binds to a database link, the user's identity is sent to the remote server.
Access controls are always evaluated on the remote server. For the user to modify or write
data successfully to the remote server, set up the correct access controls on the remote
server. For more information about how access controls are evaluated in the context of
chained operations, see Section 2.3.6, “Database Links and Access Control Evaluation”.
2.3.2. Configuring the Chaining Policy
T hese procedures describe configuring how Directory Server chains requests made by client
applications to Directory Servers that contain database links. T his chaining policy applies to all database
links created on Directory Server.
2.3.2.1. Chaining Component Operations
A component is any functional unit in the server that uses internal operations. For example, plug-ins are
considered to be components, as are functions in the front-end. However, a plug-in may actually be
comprised of multiple components (for example, the ACI plug-in).
Some components send internal LDAP requests to the server, expecting to access local data only. For
such components, control the chaining policy so that the components can complete their operations
successfully. One example is the certificate verification function. Chaining the LDAP request made by the
function to check certificates implies that the remote server is trusted. If the remote server is not trusted,
then there is a security problem.
75
Red Hat D irectory Server 8.2 Administration Guide
By default, all internal operations are not chained and no components are allowed to chain, although this
can be overridden.
Additionally, an ACI must be created on the remote server to allow the specified plug-in to perform its
operations on the remote server. T he ACI must exist in the suffix assigned to the database link.
T he following table lists component names, the potential side-effects of allowing them to chain internal
operations, and the permissions they need in the ACI on the remote server:
76
Chapter 2. Configuring D irectory D atabases
T able 2.3. Components Allowed to Chain
Component Name
Description
Permissions
ACI plug-in
T his plug-in implements access
control. Operations used to
retrieve and update ACI
attributes are not chained
because it is not safe to mix
local and remote ACI attributes.
However, requests used to
retrieve user entries may be
chained by setting the chaining
components attribute,
nsActiveChainingComponents:
cn=ACI
Plugin,cn=plugins,cn=config.
Read, search, and compare
Resource limit component
T his component sets server
limits depending on the user
bind DN. Resource limits can be
applied on remote users if the
resource limitation component is
allowed to chain. T o chain
resource limit component
operations, add the chaining
component attribute,
nsActiveChainingComponents:
cn=resource
limits,cn=components,cn=config.
Read, search, and compare
Certificate-based authentication
checking component
T his component is used when
the external bind method is
used. It retrieves the user
certificate from the database on
the remote server. Allowing this
component to chain means
certificate-based authentication
can work with a database link.
T o chain this component's
operations, add the chaining
component attribute,
nsActiveChainingComponents:
cn=certificate-based
authentication,cn=components,c
n=config.
Read, search, and compare
Password policy component
T his component is used to allow
SASL binds to the remote
server. Some forms of SASL
authentication require
authenticating with a username
and password. Enabling the
password policy allows the
server to verify and implement
the specific authentication
Read, search, and compare
77
Red Hat D irectory Server 8.2 Administration Guide
method requested and to apply
the appropriate password
policies. T o chain this
component's operations, add
the chaining component
attribute,
nsActiveChainingComponents:
cn=password
policy,cn=components,cn=config
.
SASL component
T his component is used to allow
SASL binds to the remote
server. T o chain this
component's operations, add
the chaining component
attribute,
nsActiveChainingComponents:
cn=password
policy,cn=components,cn=config
.
Read, search, and compare
Referential Integrity plug-in
T his plug-in ensures that
updates made to attributes
containing DNs are propagated
to all entries that contain
pointers to the attribute. For
example, when an entry that is a
member of a group is deleted,
the entry is automatically
removed from the group. Using
this plug-in with chaining helps
simplify the management of
static groups when the group
members are remote to the
static group definition. T o chain
this component's operations,
add the chaining component
attribute,
nsActiveChainingComponents:
cn=referential integrity
postoperation,cn=plugins,cn=co
nfig.
Read, write, search, and
compare
Attribute Uniqueness plug-in
T his plug-in checks that all the
values for a specified attribute
are unique (no duplicates). If
this plug-in is chained, it
confirms that attribute values
are unique even on attributes
changed through a database
link. T o chain this component's
operations, add the chaining
component attribute,
nsActiveChainingComponents:
cn=attribute
Read, search, and compare
78
Chapter 2. Configuring D irectory D atabases
uniqueness,cn=plugins,cn=confi
g
Roles component
T his component chains the
roles and roles assignments for
the entries in a database.
Chaining this component
maintains the roles even on
chained databases. T o chain
this component's operations,
add the chaining component
attribute,
nsActiveChainingComponents:
cn=roles,cn=components,cn=co
nfig.
Read, write, search, and
compare
NOTE
T he following components cannot be chained:
Roles plug-in
Password policy component
Replication plug-ins
Referential Integrity plug-in
When enabling the Referential Integrity plug-in on servers issuing chaining requests, be sure to
analyze performance, resource, and time needs as well as integrity needs. Integrity checks can
be time-consuming and draining on memory and CPU. For further information on the limitations
surrounding ACIs and chaining, see Section 12.1.4, “ACI Limitations”.
2.3.2.1.1. Chaining Component Operations Using the Console
1. In the Directory Server Console, select the Configuration tab.
2. Expand Data in the left pane, and click Database Link Settings.
3. Select the Settings tab in the right window.
79
Red Hat D irectory Server 8.2 Administration Guide
4. Click the Add button in the Com ponents allowed to chain section.
5. Select the component to chain from the list, and click OK.
6. Restart the server in order for the change to take effect.
After allowing the component to chain, create an ACI in the suffix on the remote server to which the
operation will be chained. For example, this creates an ACI for the Referential Integrity plug-in:
aci: (targetattr "*")(target="ldap:///ou=customers,l=us,dc=example,dc=com")
(version 3.0; acl "RefInt Access for chaining"; allow
(read,write,search,compare) userdn = "ldap:///cn=referential integrity
postoperation,cn=plugins,cn=config";)
2.3.2.1.2. Chaining Component Operations from the Command Line
1. Specify components to include in chaining using the nsActiveChainingComponents attribute in
the cn=config,cn=chaining database,cn=plugins,cn=config entry of the
configuration file.
For example, to allow the referential integrity component to chain operations, add the following to
the database link configuration file:
nsActiveChainingComponents: cn=referential integrity
postoperation,cn=components,cn=config
See T able 2.3, “Components Allowed to Chain” for a list of the components which can be chained.
2. Restart the server for the change to take effect.
service dirsrv restart instance
3. Create an ACI in the suffix on the remote server to which the operation will be chained. For
example, this creates an ACI for the Referential Integrity plug-in:
aci: (targetattr "*")(target="ldap:///ou=customers,l=us,dc=example,dc=com")
(version 3.0; acl "RefInt Access for chaining"; allow
(read,write,search,compare) userdn = "ldap:///cn=referential
integrity postoperation,cn=plugins,cn=config";)
2.3.2.2. Chaining LDAP Controls
It is possible to not chain operation requests made by LDAP controls. By default, requests made by the
following controls are forwarded to the remote server by the database link:
Virtual List View (VLV). T his control provides lists of parts of entries rather than returning all entry
80
Chapter 2. Configuring D irectory D atabases
information.
Server-side sorting. T his control sorts entries according to their attribute values, usually using a
specific matching rule.
Simple paged results. T his control breaks entry results into smaller pages, where the user specifies
the page size. T his is similar to VLV indexes for results, only without maintaining an index.
Dereferencing. T his control tracks back over references in entry attributes in a search and pulls
specified attribute information from the referenced entry and returns it with the rest of the search
results.
Managed DSA. T his controls returns smart referrals as entries, rather than following the referral, so
the smart referral itself can be changed or deleted.
Loop detection. T his control keeps track of the number of times the server chains with another
server. When the count reaches the configured number, a loop is detected, and the client application
is notified. For more information about using this control, see Section 2.4.4, “Detecting Loops”.
NOTE
Server-side sorting and VLV controls are supported only when a client application request is
made to a single database. Database links cannot support these controls when a client
application makes a request to multiple databases.
2.3.2.2.1. Chaining LDAP Controls Using the Console
1. In the Directory Server Console, select the Configuration tab.
2. Expand the Data folder in the left pane, and click Database Link Settings.
3. Select the Settings tab in the right window.
4. Click the Add button in the LDAP Controls forwarded by the database link section to
add an LDAP control to the list.
5. Select the OID of a control to add to the list, and click OK.
81
Red Hat D irectory Server 8.2 Administration Guide
2.3.2.2.2. Chaining LDAP Controls from the Command Line
T o chain controls, alter the controls that the database link forwards by changing the
nsTransmittedControls attribute of the cn=config,cn=chaining
database,cn=plugins,cn=config entry. For example, to forward the virtual list view control, add
the following to the database link entry in the configuration file:
nsTransmittedControls: 2.16.840.1.113730.3.4.9
In addition, if clients of the Directory Server create their own controls and their operations should be
chained to remote servers, add the OID of the custom control to the nsTransmittedControls attribute.
T he LDAP controls which can be chained and their OIDs are listed in the following table:
T able 2.4 . LDAP Controls and T heir OIDs
Control Name
OID
Virtual list view (VLV)
2.16.840.1.113730.3.4.9
Server-side sorting
1.2.840.113556.1.4.473
Simple paged results
1.2.840.113556.1.4.319
Managed DSA
2.16.840.1.113730.3.4.2
Loop detection
1.3.6.1.4.1.1466.29539.12
Dereferencing searches
1.3.6.1.4.1.4203.666.5.16
For more information about LDAP controls, refer to the LDAP C-SDK documentation at
http://www.mozilla.org/directory.
2.3.3. Maintaining Database Links
All of the information for the database link for the connection to the remote server.
1. In the Directory Server Console, select the Configuration tab.
2. In the left pane, expand the Data folder, and select the database link under the suffix.
3. In the right navigation pane, click the Authentication tab.
82
Chapter 2. Configuring D irectory D atabases
4. Change the connection information.
T he LDAP URL for the remote server. [2]
T he bind DN and password used by the database link to bind to the remote server.
2.3.4. Configuring Database Link Defaults
Configuring the default settings for database links defines the settings used for cascading chaining (the
number of hops allowed for a client request), the connection rules for the remote server, and how the
server responds to client requests.
1. Select the Configuration tab.
2. Expand the Data folder in the left pane, and click Database Link Settings. Open the
Default Creation Param eters tab.
3. Fill in the new configuration parameters.
83
Red Hat D irectory Server 8.2 Administration Guide
NOTE
Changes made to the default settings of a database link are not applied retroactively. Only the
database links created after changes are made to the default settings will reflect the changes.
2.3.5. Deleting Database Links
T o delete a database link, right-click the database link, and select Delete from the pop-up menu.
Confirm the delete when prompted.
1. In the Directory Server Console, select the Configuration tab.
2. Under Data in the left navigation pane, open the suffix and select the database link to delete.
3. Right-click the database link, and select Delete from the menu.
2.3.6. Database Links and Access Control Evaluation
When a user binds to a server containing a database link, the database link sends the user's identity to
the remote server. Access controls are always evaluated on the remote server. Every LDAP operation
evaluated on the remote server uses the original identity of the client application passed via the proxied
authorization control. Operations succeed on the remote server only if the user has the correct access
84
Chapter 2. Configuring D irectory D atabases
authorization control. Operations succeed on the remote server only if the user has the correct access
controls on the subtree contained on the remote server. T his requires adding the usual access controls
to the remote server with a few restrictions:
Not all types of access control can be used.
For example, role-based or filter-based ACIs need access to the user entry. Because the data are
accessed through database links, only the data in the proxy control can be verified. Consider
designing the directory in a way that ensures the user entry is located in the same database as the
user's data.
All access controls based on the IP address or DNS domain of the client may not work since the
original domain of the client is lost during chaining. T he remote server views the client application as
being at the same IP address and in the same DNS domain as the database link.
T he following restrictions apply to the ACIs used with database links:
ACIs must be located with any groups they use. If the groups are dynamic, all users in the group
must be located with the ACI and the group. If the group is static, it may refer to remote users.
ACIs must be located with any role definitions they use and with any users intended to have those
roles.
ACIs that refer to values of a user's entry (for example, userattr subject rules) will work if the user
is remote.
T hough access controls are always evaluated on the remote server, they can also be evaluated on both
the server containing the database link and the remote server. T his poses several limitations:
During access control evaluation, contents of user entries are not necessarily available (for example,
if the access control is evaluated on the server containing the database link and the entry is located
on a remote server).
For performance reasons, clients cannot do remote inquiries and evaluate access controls.
T he database link does not necessarily have access to the entries being modified by the client
application.
When performing a modify operation, the database link does not have access to the full entry stored
on the remote server. If performing a delete operation, the database link is only aware of the entry's
DN. If an access control specifies a particular attribute, then a delete operation will fail when being
conducted through a database link.
NOTE
By default, access controls set on the server containing the database link are not evaluated. T o
override this default, use the nsCheckLocalACI attribute in the cn=database_link,
cn=chaining database,cn=plugins,cn=config entry. However, evaluating access
controls on the server containing the database link is not recommended except with cascading
chaining.
2.4. Configuring Cascading Chaining
T he database link can be configured to point to another database link, creating a cascading chaining
operation. A cascading chain occurs any time more than one hop is required to access all of the data in
a directory tree.
Section 2.4.1, “Overview of Cascading Chaining”
85
Red Hat D irectory Server 8.2 Administration Guide
Section 2.4.2, “Configuring Cascading Chaining Using the Console”
Section 2.4.3, “Configuring Cascading Chaining from the Command Line”
Section 2.4.4, “Detecting Loops”
Section 2.4.5, “Summary of Cascading Chaining Configuration Attributes”
Section 2.4.6, “Cascading Chaining Configuration Example”
2.4.1. Overview of Cascading Chaining
Cascading chaining occurs when more than one hop is required for the directory to process a client
application's request.
T he client application sends a modify request to Server 1. Server one contains a database link that
forwards the operation to Server 2, which contains another database link. T he database link on Server 2
forwards the operations to server three, which contains the data the clients wants to modify in a
database. T wo hops are required to access the piece of data the client want to modify.
During a normal operation request, a client binds to the server, and then any ACIs applying to that client
are evaluated. With cascading chaining, the client bind request is evaluated on Server 1, but the ACIs
applying to the client are evaluated only after the request has been chained to the destination server, in
the above example Server 2.
For example, on Server A, a directory tree is split:
T he root suffix dc=exam ple,dc=com and the ou=people and ou=groups sub suffixes are stored
86
Chapter 2. Configuring D irectory D atabases
on Server A. T he l=europe,dc=exam ple,dc=com and ou=groups suffixes are stored in on Server
B, and the ou=people branch of the l=europe,dc=exam ple,dc=com suffix is stored on Server C.
With cascading configured on servers A, B, and C, a client request targeted at the
ou=people,l=europe,dc=exam ple,dc=com entry would be routed by the directory as follows:
First, the client binds to Server A and chains to Server B using Database Link 1. T hen Server B chains to
the target database on Server C using Database Link 2 to access the data in the
ou=people,l=europe,dc=exam ple,dc=com branch. Because at least two hops are required for
the directory to service the client request, this is considered a cascading chain.
2.4.2. Configuring Cascading Chaining Using the Console
1. Select the Configuration tab. Expand the Data folder in the left pane, and select the suffix,
then the database link.
87
Red Hat D irectory Server 8.2 Administration Guide
2. Click the Lim its and Controls tab in the right navigation pane.
3. Select the Check local ACI checkbox to enable the evaluation of local ACIs on the
intermediate database links involved in the cascading chain. Selecting this checkbox may require
adding the appropriate local ACIs to the database link.
4. Enter the maximum number of times a database link can point to another database link in the
Maxim um hops field.
By default, the maximum is ten hops. After ten hops, a loop is detected by the server, and an error
is returned to the client application.
2.4.3. Configuring Cascading Chaining from the Command Line
T o configure a cascade of database links through the command line:
1. Point one database link to the URL of the server containing the intermediate database link.
T o create a cascading chain, the nsFarmServerURL attribute of one database link must contain
the URL of the server containing another database link. Suppose the database link on the server
called exam ple1.com points to a database link on the server called africa.exam ple.com .
For example, the cn=database_link, cn=chaining database,cn=plugins,cn=config
entry of the database link on Server 1 would contain the following:
nsFarmServerURL: ldap://africa.example.com:389/
2. Configure the intermediate database link or links (in the example, Server 2) to transmit the Proxy
Authorization Control.
By default, a database link does not transmit the Proxy Authorization Control. However, when one
database link contacts another, this control is used to transmit information needed by the final
destination server. T he intermediate database link needs to transmit this control. T o configure the
database link to transmit the proxy authorization control, add the following to the
cn=config,cn=chaining database,cn=plugins,cn=config entry of the intermediate
database link:
88
Chapter 2. Configuring D irectory D atabases
nsTransmittedControls: 2.16.840.1.113730.3.4.12
T he OID value represents the Proxy Authorization Control. For more information about chaining
LDAP controls, see Section 2.3.2.2, “Chaining LDAP Controls”.
3. Create a proxy administrative user ACI on all intermediate database links.
T he ACI must exist on the server that contains the intermediate database link that checks the
rights of the first database link before translating the request to another server. For example, if
Server 2 does not check the credentials of Server 1, then anyone could bind as anonym ous and
pass a proxy authorization control allowing them more administrative privileges than appropriate.
T he proxy ACI prevents this security breach.
a. Create a database, if one does not already exist, on the server containing the intermediate
database link. T his database will contain the admin user entry and the ACI. For information
about creating a database, see Section 2.2.1, “Creating Databases”.
b. Create an entry that corresponds to the administrative user in the database.
c. Create an ACI for the administrative user that targets the appropriate suffix. T his ensures
the administrator has access only to the suffix of the database link. For example:
aci: (targetattr = "*")(version 3.0; acl "Proxied authorization for
database links";
allow (proxy) userdn = "ldap:///cn=proxy admin,cn=config";)
T his ACI is like the ACI created on the remote server when configuring simple chaining.
WARNING
Carefully examine access controls when enabling chaining to avoid giving access to
restricted areas of the directory. For example, if a default proxy ACI is created on a branch,
the users that connect through the database link will be able to see all entries below the
branch. T here may be cases when not all of the subtrees should be viewed by a user. T o
avoid a security hole, create an additional ACI to restrict access to the subtree.
4. Enable local ACI evaluation on all intermediate database links.
T o confirm that the proxy administrative ACI is used, enable evaluation of local ACIs on all
intermediate database links involved in chaining. Add the following attribute to the
cn=database_link, cn=chaining database,cn=plugins,cn=config entry of each
intermediate database link:
nsCheckLocalACI: on
Setting this attribute to on in the cn=default instance config,cn=chaining
database,cn=plugins,cn=config entry means that all new database link instances will
have the nsCheckLocalACI attribute set to on in their cn=database_link, cn=chaining
database,cn=plugins,cn=config entry.
5. Create client ACIs on all intermediate database links and the final destination database.
Because local ACI evaluation is enabled, the appropriate client application ACIs must be created
on all intermediate database links, as well as the final destination database. T o do this on the
intermediate database links, first create a database that contains a suffix that represents a root
suffix of the final destination suffix.
For example, if a client request made to the c=africa,ou=people,dc=exam ple,dc=com
suffix is chained to a remote server, all intermediate database links need to contain a database
89
Red Hat D irectory Server 8.2 Administration Guide
associated with the dc=exam ple,dc=com suffix.
Add any client ACIs to this superior suffix entry. For example:
aci: (targetattr = "*")(version 3.0; acl "Client authentication for database
link users";
allow (all) userdn = "ldap:///uid=* ,cn=config";)
T his ACI allows client applications that have a uid in the cn=config entry of Server 1 to perform
any type of operation on the data below the ou=people,dc=exam ple,dc=com suffix on server
three.
2.4.4. Detecting Loops
An LDAP control included with Directory Server prevents loops. When first attempting to chain, the server
sets this control to be the maximum number of hops, or chaining connections, allowed. Each subsequent
server decrements the count. If a server receives a count of 0, it determines that a loop has been
detected and notifies the client application.
T he number of hops allowed is defined using the nsHopLimit attribute. If not specified, the default value
is 10.
T o use the control, add the following OID to the nsTransmittedControl attribute in the
cn=config,cn=chaining database,cn=plugins,cn=config entry:
nsTransmittedControl: 1.3.6.1.4.1.1466.29539.12
If the control is not present in the configuration file of each database link, loop detection will not be
implemented.
2.4.5. Summary of Cascading Chaining Configuration Attributes
T he following table describes the attributes used to configure intermediate database links in a
cascading chain:
90
Chapter 2. Configuring D irectory D atabases
T able 2.5. Cascading Chaining Configuration Attributes
Attribute
Description
nsFarmServerURL
URL of the server containing the next database
link in the cascading chain.
nsT ransmittedControls
Enter the following OIDs to the database links
involved in the cascading chain:
nsTransmittedControls:
2.16.840.1.113730.3.4.12
nsTransmittedControls:
1.3.6.1.4.1.1466.29539.12
T he first OID corresponds to the Proxy
Authorization Control. T he second OID
corresponds to the Loop Detection Control.
aci
T his attribute must contain the following ACI:
aci: (targetattr = "*")(version 3.0;
acl "Proxied
authorization for database
links";
allow (proxy) userdn =
"ldap:///cn=proxy admin,cn=config";)
nsCheckLocalACI
T o enable evaluation of local ACIs on all
database links involved in chaining, turn local ACI
evaluation on, as follows:
nsCheckLocalACI: on
2.4.6. Cascading Chaining Configuration Example
T o create a cascading chain involving three servers as in the diagram below, the chaining components
must be configured on all three servers.
91
Red Hat D irectory Server 8.2 Administration Guide
Section 2.4.6.1, “Configuring Server One”
Section 2.4.6.2, “Configuring Server T wo”
Section 2.4.6.3, “Configuring Server T hree”
2.4 .6.1. Configuring Server One
1. Run ldapm odify[2]:
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389
-h server.example.com
2. T hen specify the configuration information for the database link, DBLink1, on Server 1, as follows:
92
Chapter 2. Configuring D irectory D atabases
dn: cn=DBLink1,cn=chaining database,cn=plugins,cn=config
objectclass: top
objectclass: extensibleObject
objectclass: nsBackendInstance
nsslapd-suffix: c=africa,ou=people,dc=example,dc=com
nsfarmserverurl: ldap://africa.example.com:389/
nsmultiplexorbinddn: cn=server1 proxy admin,cn=config
nsmultiplexorcredentials: secret
cn: DBLink1
nsCheckLocalACI:off
dn: cn=c=africa\,ou=people\,dc=example\,dc=com,cn=mapping tree,cn=config
objectclass: nsMappingTree
nsslapd-state: backend
nsslapd-backend: DBLink1
nsslapd-parent-suffix: ou=people,dc=example,dc=com
cn: c=africa\,ou=people\,dc=example\,dc=com
T he first section creates the entry associated with DBLink1. T he second section creates a new
suffix, allowing the server to direct requests made to the database link to the correct server. T he
nsCheckLocalACI attribute does not need to be configured to check local ACIs, as this is only
required on the database link, DBLink2, on Server 2.
3. T o implement loop detection, to specify the OID of the loop detection control in the
nsTransmittedControl attribute stored in cn=config,cn=chaining
database,cn=plugins,cn=config entry on Server 1.
dn: cn=config,cn=chaining database,cn=plugins,cn=config
changetype: modify
add: nsTransmittedControl
nsTransmittedControl: 1.3.6.1.4.1.1466.29539.12
As the nsTransmittedControl attribute is usually configured by default with the loop detection
control OID 1.3.6.1.4 .1.14 66.29539.12 value, it is wise to check beforehand whether it
already exists. If it does exist, this step is not necessary.
2.4 .6.2. Configuring Server T wo
1. Create a proxy administrative user on Server 2. T his administrative user will be used to allow
Server 1 to bind and authenticate to Server 2. It is useful to choose a proxy administrative user
name which is specific to Server 1, as it is the proxy administrative user which will allow server
one to bind to Server 2. Create the proxy administrative user, as follows:
dn: cn=server1 proxy admin,cn=config
objectclass: person
objectclass: organizationalPerson
objectclass: inetOrgPerson
cn: server1 proxy admin
sn: server1 proxy admin
userPassword: secret
description: Entry for use by database links
93
Red Hat D irectory Server 8.2 Administration Guide
WARNING
Do not use the Directory Manager or Administrator ID user as the proxy administrative user
on the remote server. T his creates a security hole.
2. Configure the database link, DBLink2, on Server 2, using ldapm odify:
dn: cn=DBLink2,cn=chaining database,cn=plugins,cn=config
objectclass: top
objectclass: extensibleObject
objectclass: nsBackendInstance
nsslapd-suffix: l=Zanzibar,c=africa,ou=people,dc=example,dc=com
nsfarmserverurl: ldap://zanz.africa.example.com:389/
nsmultiplexorbinddn: cn=server2 proxy admin,cn=config
nsmultiplexorcredentials: secret
cn: DBLink2
nsCheckLocalACI:on
dn: cn=l=Zanzibar\,c=africa\,ou=people\,dc=example\,dc=com,cn=mapping
tree,cn=config
objectclass: top
objectclass: extensibleObject
objectclass: nsMappingTree
nsslapd-state: backend
nsslapd-backend: DBLink2
nsslapd-parent-suffix: c=africa,ou=people,dc=example,dc=com
cn: l=Zanzibar\,c=africa\,ou=people\,dc=example\,dc=com
Since database link DBLink2 is the intermediate database link in the cascading chaining
configuration, set the nsCheckLocalACI attribute to on to allow the server to check whether it
should allow the client and proxy administrative user access to the database link.
3. T he database link on Server 2 must be configured to transmit the proxy authorization control and
the loop detection control. T o implement the proxy authorization control and the loop detection
control, specify both corresponding OIDs. Add the following information to the
cn=config,cn=chaining database,cn=plugins,cn=config entry on Server 2:
dn: cn=config,cn=chaining database,cn=plugins,cn=config
changetype: modify
add: nsTransmittedControl
nsTransmittedControl: 2.16.840.1.113730.3.4.12
nsTransmittedControl: 1.3.6.1.4.1.1466.29539.12
nsT ransm ittedControl: 2.16.84 0.1.113730.3.4 .12 is the OID for the proxy
authorization control. nsT ransm ittedControl: 1.3.6.1.4 .1.14 66.29539.12 is the or
the loop detection control.
Check beforehand whether the loop detection control is already configured, and adapt the above
command accordingly.
4. Configure the ACIs. On Server 2, ensure that a suffix exists above the
l=Zanzibar,c=africa,ou=people,dc=exam ple,dc=com suffix, so that the following
actions are possible:
Add the database link suffix
Add a local proxy authorization ACI to allow Server 1 to connect using the proxy authorization
administrative user created on Server 2
94
Chapter 2. Configuring D irectory D atabases
Add a local client ACI so the client operation succeeds on Server 2, and it can be forwarded to
server three. T his local ACI is needed because local ACI checking is turned on for the
DBLink2 database link.
Both ACIs will be placed on the database that contains the
c=africa,ou=people,dc=exam ple,dc=com suffix.
NOTE
T o create these ACIs, the database corresponding to the
c=africa,ou=people,dc=exam ple,dc=com suffix must already exist to hold the
entry. T his database needs to be associated with a suffix above the suffix specified in the
nsslapd-suffix attribute of each database link. T hat is, the suffix on the final destination
server should be a sub suffix of the suffix specified on the intermediate server.
a. Add the local proxy authorization ACI to the
c=africa,ou=people,dc=exam ple,dc=com entry:
aci:(targetattr="*")(target="l=Zanzibar,c=africa,ou=people,dc=example,dc
=com")
(version 3.0; acl "Proxied authorization for database links"; allow
(proxy)
userdn = "ldap:///cn=server1 proxy admin,cn=config";)
b. T hen add the local client ACI that will allow the client operation to succeed on Server 2,
given that ACI checking is turned on. T his ACI is the same as the ACI created on the
destination server to provide access to the
l=Zanzibar,c=africa,ou=people,dc=exam ple,dc=com branch. All users within
c=us,ou=people,dc=exam ple,dc=com may need to have update access to the
entries in l=Zanzibar,c=africa,ou=people,dc=exam ple,dc=com on server
three. Create the following ACI on Server 2 on the
c=africa,ou=people,dc=exam ple,dc=com suffix to allow this:
aci:(targetattr="*")(target="l=Zanzibar,c=africa,ou=people,dc=example,dc
=com")
(version 3.0; acl "Client authorization for database links"; allow
(all)
userdn = "ldap:///uid=*,c=us,ou=people,dc=example,dc=com";)
T his ACI allows clients that have a UID in c=us,ou=people,dc=exam ple,dc=com on
Server 1 to perform any type of operation on the
l=Zanzibar,c=africa,ou=people,dc=exam ple,dc=com suffix tree on server
three. If there are users on Server 2 under a different suffix that will require additional rights
on server three, it may be necessary to add additional client ACIs on Server 2.
2.4 .6.3. Configuring Server T hree
1. Create an administrative user on server three for Server 2 to use for proxy authorization:
95
Red Hat D irectory Server 8.2 Administration Guide
dn: cn=server2 proxy admin,cn=config
objectclass: person
objectclass: organizationalPerson
objectclass: inetOrgPerson
cn: server2 proxy admin
sn: server2 proxy admin
userPassword: secret
description: Entry for use by database links
2. T hen add the same local proxy authorization ACI to server three as on Server 2. Add the following
proxy authorization ACI to the l=Zanzibar,ou=people,dc=exam ple,dc=com entry:
aci: (targetattr = "*")(version 3.0; acl "Proxied authorization
for database links"; allow (proxy) userdn = "ldap:///cn=server2
proxy admin,cn=config";)
T his ACI gives the Server 2 proxy admin read-only access to the data contained on the remote
server, server three, within the l=Zanzibar,ou=people,dc=exam ple,dc=com subtree only.
3. Create a local client ACI on the l=Zanzibar,ou=people,dc=exam ple,dc=com subtree that
corresponds to the original client application. Use the same ACI as the one created for the client
on Server 2:
aci: (targetattr
="*")(target="l=Zanzibar,c=africa,ou=people,dc=example,dc=com")
(version 3.0; acl "Client authentication for database link users"; allow
(all)
userdn = "ldap:///uid=*,c=us,ou=people,dc=example,dc=com";)
T he cascading chaining configuration is now set up. T his cascading configuration allows a user to bind
to Server 1 and modify information in the
l=Zanzibar,c=africa,ou=people,dc=exam ple,dc=com branch on server three. Depending on
your security needs, it may be necessary to provide more detailed access control.
2.5. Using Referrals
Referrals tell client applications which server to contact for a specific piece of information. T his
redirection occurs when a client application requests a directory entry that does not exist on the local
server or when a database has been taken off-line for maintenance. T his section contains the following
information about referrals:
Section 2.5.1, “Starting the Server in Referral Mode”
Section 2.5.2, “Setting Default Referrals”
Section 2.5.3, “Creating Smart Referrals”
Section 2.5.4, “Creating Suffix Referrals”
For conceptual information on how to use referrals in the directory, see the Directory Server Deployment
Guide.
2.5.1. Starting the Server in Referral Mode
Referrals are used to redirect client applications to another server while the current server is unavailable
or when the client requests information that is not held on the current server. For example, starting
Directory Server in referral mode while there are configuration changes being made to the Directory
Server will refer all clients to another supplier while that server is unavailable. Starting the Directory
96
Chapter 2. Configuring D irectory D atabases
Server in referral mode is done with the refer command.
Run nsslapd with the refer option.
/usr/sbin/ns-slapd refer -D /etc/dirsrv/slapd-instance_name [-p port] -r
referral_url
/etc/dirsrv/slapd-instance_name is the directory where the Directory Server configuration
files are. T his is the default location on Red Hat Enterprise Linux 5 (64-bit).
port is the optional port number of the Directory Server to start in referral mode.
referral_url is the referral returned to clients. T he format of an LDAP URL is covered in Appendix B,
LDAP URLs.
2.5.2. Setting Default Referrals
Default referrals are returned to client applications that submit operations on a DN not contained within
any of the suffixes maintained by the directory. T he following procedures describes setting a default
referral for the directory using the console and the command-line utilities.
2.5.2.1. Setting a Default Referral Using the Console
1. In the Directory Server Console, select the Configuration tab.
2. Select the top entry in the navigation tree in the left pane.
3. Select the Settings tab in the right pane.
4. Enter an LDAP URL for the referral.
Enter multiple referral URLs separated by spaces and in quotes:
"ldap://dir1.example.com:389/dc=example,dc=com" "ldap://dir2.example.com/"
For more information about LDAP URLs, see Appendix B, LDAP URLs.
2.5.2.2. Setting a Default Referral from the Command Line
ldapm odify can add a default referral to the cn=config entry in the directory's configuration file. For
example, to add a new default referral from one Directory Server, dir1.exam ple.com , to a server
named dir2.exam ple.com , add a new line to the cn=config entry.
1. Run the ldapm odify utility:[2]
97
Red Hat D irectory Server 8.2 Administration Guide
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
ldapm odify binds to the server and prepares it to change an entry in the configuration file.
2. Add the default referral to the dir2.exam ple.com server:
dn: cn=config
changetype: modify
replace: nsslapd-referral
nsslapd-referral: ldap://dir2.example.com/
After adding the default referral to the cn=config entry of the directory, the directory will return the
default referral in response to requests made by client applications. T he Directory Server does not need
to be restarted.
2.5.3. Creating Smart Referrals
Smart referrals map a directory entry or directory tree to a specific LDAP URL. Using smart referrals,
client applications can be referred to a specific server or a specific entry on a specific server.
For example, a client application requests the directory entry
uid=jdoe,ou=people,dc=exam ple,dc=com . A smart referral is returned to the client that points to
the entry cn=john doe,o=people,l=europe,dc=exam ple,dc=com on the server
directory.europe.exam ple.com .
T he way the directory uses smart referrals conforms to the standard specified in RFC 2251 section
4.1.11. T he RFC can be downloaded at http://www.ietf.org/rfc/rfc2251.txt.
2.5.3.1. Creating Smart Referrals Using the Directory Server Console
1. In the Directory Server Console, select the Directory tab.
2. Browse through the tree in the left navigation pane, and select the entry for which to add the
referral.
3. Right-click the entry, and select Set Sm art Referrals.
4. Select the Enable Sm art Referral checkbox. (Unchecking the option removes all smart
referrals from the entry and deletes the referral object class from the entry.)
98
Chapter 2. Configuring D irectory D atabases
5. In the Enter a new Sm art Referral field, enter a referral in the LDAP URL format, and then
click Add. T he LDAP URL must be in the following format:
ldap://hostname:portnumber/[optional_dn]
optional_dn is the explicit DN for the server to return to the requesting client application.
Construct opens a wizard to direct the process of adding a referral.
T he Sm art Referral List lists the referrals currently in place for the selected entry. T he
entire list of referrals is returned to client applications in response to a request with the Return
Referrals for All Operations or Return Referrals for Update Operations
options in the Suffix Settings tab, which is available under the Configuration tab.
T o modify the list, click Edit to edit the selected referral or Delete to delete the selected
referral.
6. T o set the referral to use different authentication credentials, click Authentication, and specify
the appropriate DN and password. T his authentication remains valid only until the Console is
closed; then it's reset to the same authentication used to log into the Console.
99
Red Hat D irectory Server 8.2 Administration Guide
2.5.3.2. Creating Smart Referrals from the Command Line
Use the ldapm odify command-line utility[2] to create smart referrals from the command line.
T o create a smart referral, create the relevant directory entry, and add the referral object class. T his
object class allows a single attribute, ref. T he ref attribute must contain an LDAP URL.
For example, add the following to return a smart referral for an existing entry, uid=jdoe:
dn: uid=jdoe,ou=people,dc=example,dc=com
objectclass: referral
ref:
ldap://directory.europe.example.com/cn=john%20doe,ou=people,l=europe,dc=example,d
c=com
NOTE
Any information after a space in an LDAP URL is ignored by the server. For this reason, use %20
instead of spaces in any LDAP URL used as a referral.
T o add the entry uid=jdoe,ou=people,dc=exam ple,dc=com with a referral to
directory.europe.exam ple.com , include the following in the LDIF file before importing:
dn: uid=jdoe,ou=people,dc=example,dc=com
objectclass: top
objectclass: person
objectclass: organizationalPerson
objectclass: inetOrgPerson
objectclass: referral
cn: john doe
sn: doe
uid: jdoe
ref:
ldap://directory.europe.example.com/cn=john%20doe,ou=people,l=europe,dc=example,d
c=com
Use the -M option with ldapm odify when there is already a referral in the DN path. For information
100
Chapter 2. Configuring D irectory D atabases
about the ldapm odify utility, see the Directory Server Configuration and Command-Line Tool
Reference.
2.5.4. Creating Suffix Referrals
T he following procedure describes creating a referral in a suffix. T his means that the suffix processes
operations using a referral rather than a database or database link.
WARNING
When a suffix is configured to return referrals, the ACIs contained by the database associated
with the suffix are ignored.
2.5.4 .1. Creating Suffix Referrals Using the Console
Referrals can be used to point a client application temporarily to a different server. For example, adding a
referral to a suffix so that the suffix points to a different server allows the database associated with the
suffix is taken off-line for maintenance without affecting the users of the Directory Server database.
T o set referrals in a suffix:
1. In the Directory Server Console, select the Configuration tab.
2. Under Data in the left pane, select the suffix for which to add a referral.
3. Click the Suffix Settings tab, and select the Return Referrals for ... Operations
radio button.
Selecting Return Referrals for Update Operations means that the directory redirects
only update and write requests to a read-only database. For example, there may be a local copy of
directory data, and that data should be available for searches but not for updates, so it's
replicated across several servers. Enabling referrals for that Directory Server only for update
requests means that when a client asks to update an entry, the client is referred to the server that
owns the data, where the modification request can proceed.
4. Click the Referrals tab. Enter an LDAP URL in the
click Construct to create an LDAP URL.
[3]
in the Enter a new referral field, or
101
Red Hat D irectory Server 8.2 Administration Guide
5. Click Add to add the referral to the list.
You can enter multiple referrals. T he directory returns the entire list of referrals in response to
requests from client applications.
2.5.4 .2. Creating Suffix Referrals from the Command Line
Add a suffix referral to the root or sub suffix entry in the directory configuration file under the
cn=m apping tree,cn=config branch.
1. Run ldapm odify.[2] For example:
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389
-h server.example.com
T he ldapm odify utility binds to the server and prepares it to add information to the
configuration file.
2. Add a suffix referral to the ou=people,dc=exam ple,dc=com root suffix, as follows:
dn: cn=ou=people,dc=example,dc=com,cn=mapping tree,cn=config
objectclass: extensibleObject
objectclass: nsMappingTree
nsslapd-state: referral
nsslapd-referral: ldap://zanzibar.com/
T he nsslapd-state attribute is set to referral, meaning that a referral is returned for
requests made to this suffix. T he nsslapd-referral attribute contains the LDAP URL of the
referral returned by the suffix, in this case a referral to the zanzibar.com server.
T he nsslapd-state attribute can also be set to referral on update. T his means that the
database is used for all operations except update requests. When a client application makes an
update request to a suffix set to referral on update, the client receives a referral.
102
Chapter 2. Configuring D irectory D atabases
For more information about the suffix configuration attributes, refer to T able 2.1, “Suffix Attributes”.
[2] The LDAP to o ls referenc ed in this g uid e are Mo z illa LDAP, ins talled with Direc to ry Server in the /usr/l i b 6 4 /m o zl d ap
d irec to ry o n Red Hat Enterp ris e Linux 5 (6 4-b it); d irec to ries fo r o ther p latfo rms are lis ted in Sec tio n 1.3, “ LDAP To o l Lo c atio ns ” .
Ho wever, Red Hat Enterp ris e Linux s ys tems als o inc lud e LDAP to o ls fro m O p enLDAP. It is p o s s ib le to us e the O p enLDAP
c o mmand s as s ho wn in the examp les , b ut yo u mus t us e the - x arg ument to d is ab le SASL and allo w s imp le authentic atio n.
[2] Unlike the s tand ard LDAP URL fo rmat, the URL o f the remo te s erver d o es no t s p ec ify a s uffix. It has the fo rm
l d ap ://hostname:port/.
[3] Ap p end ix B, LDAP URLs has mo re info rmatio n ab o ut the s truc ture o f LDAP URLs .
103
Red Hat D irectory Server 8.2 Administration Guide
Chapter 3. Creating Directory Entries
T his chapter discusses how to use the Directory Server Console and the ldapm odify and
ldapdelete command-line utilities to modify the contents of your directory.
Entries stored in Active Directory can be added to the Directory Server through Windows Sync; see
Chapter 10, Synchronizing Red Hat Directory Server with Microsoft Active Directory for more information
on adding or modifying synchronized entries through Windows User Sync.
3.1. Managing Entries from the Directory Console
You can use the Directory tab and the Property Editor on the Directory Server Console to add,
modify, or delete entries individually.
T o add several entries simultaneously, use the command-line utilities described in Section 3.2,
“Managing Entries from the Command Line”.
Section 3.1.1, “Creating a Root Entry”
Section 3.1.2, “Creating Directory Entries”
Section 3.1.3, “Modifying Directory Entries”
Section 3.1.4, “Deleting Directory Entries”
NOTE
You cannot modify your directory unless the appropriate access control rules have been set. For
information on creating access control rules for your directory, see Chapter 12, Managing Access
Control.
3.1.1. Creating a Root Entry
Each time a new database is created, it is associated with the suffix that will be stored in the database.
T he directory entry representing that suffix is not automatically created.
T o create a root entry for a database:
1. In the Directory Server Console, select the Configuration tab.
2. Right-click on the Data entry in the left menu, and select New Root Suffix from the menu.
104
Chapter 3. Creating D irectory Entries
3. Fill in the new suffix and database information.
4. In the Directory tab, right-click the top object representing the Directory Server, and choose
New Root Object.
T he secondary menu under New Root Object displays the new suffixes without a
corresponding directory entry. Choose the suffix corresponding to the entry to create.
5. In the New Object window, select the object class corresponding to the new entry.
105
Red Hat D irectory Server 8.2 Administration Guide
T he object class must contain the attribute used to name the suffix. For example, if the entry
corresponds to the suffix ou=people,dc=exam ple,dc=com , then choose the
organizationalUnit object class or another object class that allows the ou attribute.
6. Click OK in the New Object window.
T he Property Editor for the new entry opens. You can either accept the current values by clicking
OK or modify the entry, as explained in Section 3.1.3, “Modifying Directory Entries”.
3.1.2. Creating Directory Entries
Directory Server Console offers predefined templates, with preset forms, for new directory entries.
T able 3.1, “Entry T emplates and Corresponding Object Classes” shows what type of object class is
used for each template.
T able 3.1. Entry T emplates and Corresponding Object Classes
T emplate
Object Class
User
inetOrgPerson
Group
groupOfUniqueNames
Organizational Unit
organizationalUnit
Role
nsRoleDefinition
Class of Service
cosSuperDefinition
Another type, Other allows any kind of entry to be created by allowing users to select the specific object
classes and attributes to apply.
1. In the Directory Server Console, select the Directory tab.
2. In the left pane, right-click the main entry to add the new entry, and select the type of entry: User,
Group, Organizational Unit, Role, Class of Service, or Other.
106
Chapter 3. Creating D irectory Entries
3. If the new entry type was Other, then a list of object classes opens. Select an object class from
the list to define the new entry.
4. Supply a value for all the listed attributes. Required attributes are marked with an asterisk (* ).
5. T o display the full list of attributes available for the object class (entry type), click the Advanced
107
Red Hat D irectory Server 8.2 Administration Guide
button.
In the Property Editor, select any additional attributes, and fill in the attribute values.
6. Click OK to save the entry. T he new entry is listed in the right pane.
3.1.3. Modifying Directory Entries
Modifying directory entries in Directory Server Console uses a dialog window called the Property
Editor. T he Property Editor contains the list of object classes and attributes belonging to an
entry and can be used to edit the object classes and attributes belonging to that entry by adding and
removing object classes, attributes and attribute values, and attribute subtypes.
108
Chapter 3. Creating D irectory Entries
T he Property Editor can be opened in several ways:
From the Directory tab, by right-clicking an entry, and selecting Advanced Properties from the
pop-up menu.
From the Directory tab, by double-clicking an entry and clicking the Advanced button
From the Create... new entry forms, by clicking the Advanced button
From the New Object window, by clicking OK
3.1.3.1. Adding or Removing an Object Class to an Entry
T o add an object class to an entry:
1. In the Directory tab of the Directory Server Console, right-click the entry to modify, and select
Advanced from the pop-up menu.
109
Red Hat D irectory Server 8.2 Administration Guide
2. Select the object class field, and click Add Value.
T he Add Object Class window opens. It shows a list of object classes that can be added to
the entry.
3. Select the object class to add, and click OK.
110
Chapter 3. Creating D irectory Entries
T o remove an object class from an entry, click the text box for the object class to remove, and then click
Delete Value.
3.1.3.2. Adding an Attribute to an Entry
Before you can add an attribute to an entry, the entry must contain an object class that either requires or
allows the attribute. See Section 3.1.3.1, “Adding or Removing an Object Class to an Entry” and
Chapter 6, Managing the Directory Schema for more information.
T o add an attribute to an entry:
1. In the Directory tab of the Directory Server Console, right-click the entry to modify, and select
Advanced from the pop-up menu.
2. Click Add Attribute.
111
Red Hat D irectory Server 8.2 Administration Guide
3. Select the attribute to add from the list, and click OK.
NOTE
If the attribute you want to add is not listed, add the object class containing the attribute
first, then add the attribute. See Section 3.1.3.1, “Adding or Removing an Object Class to an
Entry” for instructions on adding an object class. If you do not know which object class
contains the attribute you need, look up the attribute in the Directory Server Schema
Reference, which lists the object classes which use that attribute.
4. T ype in the value for the new attribute in the field to the right of the attribute name.
T o remove the attribute and all its values from the entry, select Delete Attribute from the Edit
menu.
112
Chapter 3. Creating D irectory Entries
3.1.3.3. Adding Very Large Attributes
T he configuration attribute nsslapd-maxbersize sets the maximum size limit for LDAP requests. T he
default configuration of Directory Server sets this attribute at 2 megabytes. LDAP add or modify
operations will fail when attempting to add very large attributes that result in a request that is larger than
2 megabytes.
T o add very large attributes, first change the setting for the nsslapd-maxbersize configuration
attribute to a value larger than the largest LDAP request you will make.
When determining the value to set, consider all elements of the LDAP add and modify operations used to
add the attributes, not just the single attribute. T here are a number of different factors to consider,
including the following:
T he size of each attribute name in the request
T he size of the values of each of the attributes in the request
T he size of the DN in the request
Some overhead, usually 10 kilobytes
One common issue that requires increasing the nsslapd-maxbersize setting is using attributes which
hold CRL values, such as certificateRevocationList, authorityRevocationList, and
deltaRevocationList.
For further information about the nsslapd-maxbersize attribute and for information about setting this
attribute, see the section "nsslapd-maxbersize (MaximumMessage Size)" in chapter 2, "Core Server
Configuration Reference," in Directory Server Configuration and Command-Line Tool Reference.
3.1.3.4 . Adding Attribute Values
Multi-valued attributes allow multiple value for one attribute to be added to an entry.
1. In the Directory tab of the Directory Server Console, right-click the entry to modify, and select
Advanced from the pop-up menu.
2. Select the attribute to which to add a value, and then click Add Value.
3. T ype in the new attribute value.
113
Red Hat D irectory Server 8.2 Administration Guide
T o remove an attribute value from an entry, click the text box of the attribute value to remove, and click
Delete Value.
3.1.3.5. Adding an Attribute Subtype
A subtype allows the same entry value to be represented in different ways, such as providing a foreigncharacterset version. T here three different kinds of subtypes to attributes which can be added to an
entry: language, binary, and pronunciation.
T o add a subtype to an entry:
1. In the Directory tab of the Directory Server Console, right-click the entry to modify, and select
Properties from the pop-up menu.
2. Click Add Attribute, and select the attribute to add from the list.
3. Add a language subtype by selecting a value from the Language drop-down list. Add either a
binary or pronunciation subtypeby selecting a value from the Subtype drop-down list.
Language Subtype
Sometimes a user's name can be more accurately represented in characters of a language other than
the default language. For example, a user, Noriko, has a name in Japanese and prefers that her name
be represented by Japanese characters when possible. You can select Japanese as a language
subtype for the givennam e attribute so that other users can search for her name in Japanese as well
as English. For example:
114
Chapter 3. Creating D irectory Entries
givenname;lang-ja
T o specify a language subtype for an attribute, add the subtype to the attribute name as follows:
attribute;lang-subtype:attribute value
attribute is the attribute being added to the entry and subtype is the two character abbreviation for the
language. T he supported language subtypes are listed in T able C.2, “Supported Language Subtypes”.
Only one language subtype can be added per attribute instance in an entry. T o assign multiple language
subtypes, add another attribute instance to the entry, and then assign the new language subtype. For
example, the following is illegal:
cn;lang-ja;lang-en-GB:value
Instead, use:
cn;lang-ja:ja-value
cn;lang-en-GB:value
Binary Subtype
Assigning the binary subtype to an attribute indicates that the attribute value is binary, such as user
certificates (usercertificate;binary).
Although you can store binary data within an attribute that does not contain the binary subtype (for
example, jpegphoto), the binary subtype indicates to clients that multiple variants of the attribute
type may exist.
Pronunciation Subtype
Assigning the pronunciation subtype to an attribute indicates that the attribute value is a phonetic
representation. T he subtype is added to the attribute name as attribute;phonetic. T his subtype is
commonly used in combination with a language subtype for languages that have more than one
alphabet, where one is a phonetic representation.
T his subtype is useful with attributes that are expected to contain user names, such as cn or
givennam e. For example, givennam e;lang-ja;phonetic indicates that the attribute value is the
phonetic version of the user's Japanese name.
3.1.4. Deleting Directory Entries
1. In the Directory Server Console, select the Directory tab.
2. Right-click the entry to delete, and select Delete from the right-click menu.
115
Red Hat D irectory Server 8.2 Administration Guide
WARNING
T he server deletes the entry or entries immediately. T here is no way to undo the delete
operation.
3.2. Managing Entries from the Command Line
T he command-line utilities allow you to manipulate the contents of your directory. T hey can be useful to
write scripts to perform bulk management of the directory or to test the Directory Server. For example,
you might want to ensure that it returns the expected information after you have made changes to
access control information.
With command-line utilities, information can be provided directly from the command line or through an
LDIF input file.
Section 3.2.1, “Providing Input from the Command Line”
Section 3.2.2, “Creating a Root Entry from the Command Line”
Section 3.2.3, “Adding Entries Using LDIF”
Section 3.2.4, “Adding and Modifying Entries Using ldapmodify”
Section 3.2.5, “Deleting Entries Using ldapdelete”
Section 3.2.6, “Using Special Characters”
116
Chapter 3. Creating D irectory Entries
NOTE
You cannot modify your directory unless the appropriate access control rules have been set. For
information on creating access control rules for the directory, see Chapter 12, Managing Access
Control.
3.2.1. Providing Input from the Command Line
When you provide input to the ldapm odify and ldapdelete [4] utilities directly from the command
line, you must use LDIF statements. For detailed information on LDIF statements, see Section 3.3, “Using
LDIF Update Statements to Create or Modify Entries”.
T he ldapm odify and ldapdelete utilities read the statements that you enter in exactly the same
way as if they were read from a file. When all of the input has been entered, enter the character that the
shell recognizes as the end of file (EOF) escape sequence. T he utility then begins operations based on
the supplied inputs.
While the EOF escape sequence depends on the type of machine, the EOF escape sequence almost
always control-D (^D).
For example, to input some LDIF update statements to ldapm odify:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=Barry Nixon,ou=people,dc=example,dc=com
changetype: modify
delete: telephonenumber
add: manager
manager: cn=Harry Cruise,ou=people,dc=example,dc=com
^D
When adding an entry from the command line or from LDIF, make sure that an entry representing a
subtree is created before new entries are created under that branch. For example, to place an entry in a
People subtree, create an entry representing that subtree before creating entries within the subtree.
For example:
dn: dc=example,dc=com
dn: ou=People,dc=example,dc=com
...People subtree entries. ...
dn: ou=Group,dc=example,dc=com
...Group subtree entries. ...
3.2.2. Creating a Root Entry from the Command Line
T he ldapm odify command-line utility can be used to create a new root entry in a database. For
example:
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389 -h
server.example.com
T he ldapm odify utility binds to the server and prepares it to add an entry. T he new root object can
then be added, as follows:
117
Red Hat D irectory Server 8.2 Administration Guide
dn: Suffix_Name
objectclass: newobjectclass
T he DN corresponds to the DN of the root or sub-suffix contained by the database. T he newobjectclass
value depends upon the type of object class you are adding to the database. You may need to specify
additional required attributes depending on the type of root object being added.
NOTE
You can use ldapm odify to add root objects only if you have one database per suffix. If you
create a suffix that is stored in several databases, you must use the ldif2db utility with the noption parameter to specify the database that will hold the new entries. For information, see
Section 4.1.4, “Importing from the Command Line”.
3.2.3. Adding Entries Using LDIF
You can use an LDIF file to add multiple entries or to import an entire database. T o add entries using an
LDIF file and the Directory Server Console:
1. Define the entries in an LDIF file.
LDIF files are described in Appendix A, LDAP Data Interchange Format.
2. Import the LDIF file from the Directory Server Console.
See Section 4.1.2, “Importing a Database from the Console” for information about LDIF file formats.
When you import the LDIF file, select Append to database in the Im port dialog box so that
the server will only import entries that do not currently exist in the directory.
You can also add entries described in an LDIF file from the command line using the ldapm odify
command with the -f option.
3.2.4. Adding and Modifying Entries Using ldapmodify
T he ldapm odify[4] command can add and modify entries in an existing Directory Server database.
T he ldapm odify command opens a connection to the specified server using the supplied
distinguished name and password and modifies the entries based on LDIF update statements contained
in a specified file. Because ldapm odify uses LDIF update statements, ldapm odify can do
everything that ldapdelete can do.
Consider the following when using ldapm odify:
If the server detects an attribute or object class in the entry that is not known to the server, then the
modify operation will fail when it reaches the erroneous entry. All entries that were processed before
the error was encountered will be successfully added or modified. If you run ldapm odify with the c option (do not stop on errors), all correct entries processed after the erroneous entry will be
successfully added or modified.
If a required attribute is not present, the modify operation fails. T his happens even if the offending
object class or attribute is not being modified.
118
Chapter 3. Creating D irectory Entries
NOTE
T o create the root entry a database suffix (such as dc=exam ple,dc=com ) using
ldapm odify, you must bind to the directory as the Directory Manager.
3.2.4 .1. Adding Entries Using ldapmodify
T ypically, to add the entries using ldapm odify, specify the DN and password to bind to the Directory
Server, the port and host of the Directory Server, and the LDIF file to use. For example:
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389 -h
server.example.com -f new.ldif
T his ldapm odify example has the following values:
T he entries to be created are specified in the file new.ldif. (In this example, the LDIF statements in
the new.ldif file do not specify a change type. T hey follow the format defined in Section A.1, “About
the LDIF File Format”.)
T he Directory Manager is a database administrator who has the authority to modify the entries, and
its password is secret.
T he hostname is exam ple.com .
T he server uses port number 389.
T able 3.2, “ldapmodify Parameters Used for Adding Entries” describes the ldapm odify parameters
used in the example.
T able 3.2. ldapmodify Parameters Used for Adding Entries
Parameter Name
Description
-a
Specifies that the modify operation will add new
entries to the directory.
-D
Specifies the distinguished name with which to
authenticate to the server. T he value must be a
DN recognized by the Directory Server, and it
must also have the authority to modify the entries.
-w
Specifies the password associated with the
distinguished name specified in the -D parameter.
-h
Specifies the name of the host on which the
server is running.
-p
Specifies the port number that the server uses.
-f
Optional parameter that specifies the file
containing the LDIF update statements used to
define the modifications. If you do not supply this
parameter, the update statements are read from
stdin. For information on supplying LDIF update
statements from the command line, see
Section 3.2.1, “Providing Input from the Command
Line”.
119
Red Hat D irectory Server 8.2 Administration Guide
For full information on ldapm odify parameters, see the Directory Server Configuration and CommandLine Tool Reference.
3.2.4 .2. Modifying Entries Using ldapmodify
T ypically, to edit entries using ldapm odify, specify the DN and password to bind to the Directory
Server, the port and host of the Directory Server, and the LDIF file to use, as when adding entries with
ldapm odify. For example:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com -f modify_statements
T his ldapm odify example has the following values:
T he entries to modify are specified in the file m odify_statem ents. Before the entries can be
modified, you must first create the m odify_statem ents file with the appropriate LDIF update
statements; LDIF update statements are described in Section 3.3, “Using LDIF Update Statements to
Create or Modify Entries”.
T he bind DN is cn=Directory Manager, which has permissions to edit any entry in the database,
and the password is secret.
T he hostname is exam ple.com .
T he server uses port number 389.
T able 3.3, “ldapmodify Parameters Used for Modifying Entries” describes the ldapm odify parameters
used in the example.
T able 3.3. ldapmodify Parameters Used for Modifying Entries
Parameter Name
Description
-D
Specifies the distinguished name with which to
authenticate to the server. T he value must be a
DN recognized by the Directory Server, and it
must also have the authority to modify the entries.
-w
Specifies the password associated with the
distinguished name specified in the -D parameter.
-h
Specifies the name of the host on which the
server is running.
-p
Specifies the port number that the server uses.
-f
Optional parameter that specifies the file
containing the LDIF update statements used to
define the modifications. If you do not supply this
parameter, the update statements are read from
stdin. For information on supplying LDIF update
statements from the command line, see
Section 3.2.1, “Providing Input from the Command
Line”.
For full information on ldapm odify parameters, see the Directory Server Configuration and CommandLine Tool Reference.
3.2.5. Deleting Entries Using ldapdelete
120
Chapter 3. Creating D irectory Entries
T he ldapdelete command-line utility opens a connection to the specified server using the provided
distinguished name and password and deletes the specified entry or entries.
NOTE
You can only delete entries at the end of a branch. You cannot delete entries that are branch
points in the directory tree.
For example, of the following three entries, only the last two entries can be deleted.
ou=People,dc=example,dc=com
cn=Paula Simon,ou=People,dc=example,dc=com
cn=Jerry O'Connor,ou=People,dc=example,dc=com
T he entry that identifies the People subtree can be deleted only if there are not any entries below it. T o
delete ou=People,dc=exam ple,dc=com , you must first delete Paula Simon and Jerry O'Connor's
entries and all other entries in that subtree.
Like ldapm odify, running ldapdelete requires the DN and password to bind to the Directory Server,
the port and host of the Directory Server, and the DNs of the entries to delete. For example:
ldapdelete -D "cn=Directory Manager" -w secret -h example.com -p 389 "cn=Robert
Jenkins,ou=People,dc=example,dc=com" "cn=Lisa
Jangles,ou=People,dc=example,dc=com"
T his ldapdelete example has the following values:
T he entries to delete have the DNs cn=Robert Jenkins,ou=People,dc=exam ple,dc=com
and cn=Lisa Jangles,ou=People,dc=exam ple,dc=com .
T he bind DN is the Directory Manager, which has permission to delete every entry in the database,
and a password of secret.
T he hostname is exam ple.com .
T he server uses port number 389.
T able 3.4, “ldapdelete Parameters Used for Deleting Entries” describes the ldapdelete parameters
used in the example:
T able 3.4 . ldapdelete Parameters Used for Deleting Entries
Parameter Name
Description
-D
Specifies the distinguished name with which to
authenticate to the server. T he value must be a
DN recognized by the Directory Server, and it
must also have the authority to modify the entries.
-w
Specifies the password associated with the
distinguished name specified in the -D parameter.
-h
Specifies the name of the host on which the
server is running.
-p
Specifies the port number that the server uses.
121
Red Hat D irectory Server 8.2 Administration Guide
For full information on ldapdelete parameters, see the Directory Server Configuration and CommandLine Tool Reference.
3.2.6. Using Special Characters
When using the Directory Server command-line client tools, you may need to specify values that contain
characters that have special meaning to the command-line interpreter, such as space ( ), asterisk (* ), or
backslash (\). When this situation occurs, enclose the value in quotation marks (""). For example:
-D "cn=Barbara Jensen,ou=Product Development,dc=example,dc=com"
Depending on the command-line utility, use either single or double quotation marks; see your operating
system documentation for more information.
Additionally, if a DN contains commas, you must escape the commas with a backslash (\). For example:
-D "cn=Patricia Fuentes,ou=people,o=example.com Bolivia\,S.A."
T o delete user Patricia Fuentes from the exam ple.com Bolivia, S.A. tree, use the following
command:
ldapdelete -D "cn=Directory Manager" -w secret -h example.com -p 389 "cn=Patricia
Fuentes,ou=People,o=example.com Bolivia\,S.A."
3.3. Using LDIF Update Statements to Create or Modify Entries
LDIF update statements define how ldapm odify changes the directory entry. In general, LDIF update
statements contain the following information:
T he DN of the entry to be modified.
A changetype that defines how a specific entry is to be modified (add, delete, m odify, m odrdn).
A series of attributes and their changed values.
A change type is required unless ldapm odify is run with the -a parameter. If you specify the -a
parameter, then an add operation (changetype: add) is assumed. However, any other change type
overrides the -a parameter.
If you specify a modify operation (changetype: m odify), a change operation is required that
indicates how the entry should be changed.
If you specify changetype: m odrdn, change operations are required that specify how the relative
distinguished name (RDN) is to be modified. A distinguished name's RDN is the left-most value in the
DN. For example, the distinguished name uid=ssarette,dc=exam ple,dc=com has an RDN of
uid=ssarette.
T he general format of LDIF update statements is as follows:
dn: distinguished_name
changetype: changetype_identifier
change_operation_identifier: list_of_attributes
change_operation_identifier: list_of_attributes
A dash (-) must be used to denote the end of a change operation if subsequent change operations are
122
Chapter 3. Creating D irectory Entries
specified. For example, the following statement adds the telephone number and manager attributes to
the entry:
dn: cn=Lisa Jangles,ou=People,dc=example,dc=com
changetype: modify
add: telephonenumber
telephonenumber: (408) 555-2468
add: manager
manager: cn=Harry Cruise,ou=People,dc=example,dc=com
In addition, the line continuation operator is a single space. T herefore, the following two statements are
identical:
dn: cn=Lisa Jangles,ou=People,dc=example,dc=com
dn: cn=Lisa Jangles,
ou=People,
dc=example,dc=com
T he following sections describe the change types in detail.
3.3.1. Adding an Entry Using LDIF
changetype: add adds an entry to the directory. When you add an entry, make sure to create an
entry representing a branch point before you try to create new entries under that branch. T hat is, to
place an entry in a People and a Groups subtree, then create the branch point for those subtrees
before creating entries within the subtrees. For example:
123
Red Hat D irectory Server 8.2 Administration Guide
dn: dc=example,dc=com
changetype: add
objectclass: top
objectclass: organization
o: example.com
dn: ou=People,dc=example,dc=com
changetype: add
objectclass: top
objectclass: organizationalUnit
ou: People
ou: Marketing
dn: cn=Pete Minsky,ou=People,dc=example,dc=com
changetype: add
objectclass: top
objectclass: person
objectclass: organizationalPerson
objectclass: inetOrgPerson
cn: Pete Minsky
givenName: Pete
sn: Minsky
ou: People
ou: Marketing
uid: pminsky
dn: cn=Sue Jacobs,ou=People,dc=example,dc=com
changetype: add
objectclass: top
objectclass: person
objectclass: organizationalPerson
objectclass: inetOrgPerson
cn: Sue Jacobs
givenName: Sue
sn: Jacobs
ou: People
ou: Marketing
uid: sjacobs
dn: ou=Groups,dc=example,dc=com
changetype: add
objectclass: top
objectclass: organizationalUnit
ou: Groups
dn: cn=Administrators,ou=Groups,dc=example,dc=com
changetype: add
objectclass: top
objectclass: groupOfNames
member: cn=Sue Jacobs,ou=People,dc=example,dc=com
member: cn=Pete Minsky,ou=People,dc=example,dc=com
cn: Administrators
dn: ou=example.com Bolivia\, S.A.,dc=example,dc=com
changetype: add
objectclass: top
objectclass: organizationalUnit
ou: example.com Bolivia\, S.A.
dn: cn=Carla Flores,ou=example.com Bolivia\,S.A.,dc=example,dc=com
124
Chapter 3. Creating D irectory Entries
changetype: add
objectclass: top
objectclass: person
objectclass: organizationalPerson
objectclass: inetOrgPerson
cn: Carla Flores
givenName: Carla
sn: Flores
ou: example.com Bolivia\, S.A.
uid: cflores
3.3.2. Renaming an Entry Using LDIF
changetype: m odrdn changes an entry's relative distinguished name (RDN). An entry's RDN is the
left-most element in the distinguished name. T he RDN for cn=Barry
Nixon,ou=People,dc=exam ple,dc=com is cn=Barry Nixon, and the RDN for
ou=People,dc=exam ple,dc=com is ou=People. A changetype: m odrdn operation changes
that left-most value in an entry's DN.
T he m odrdn change type only changes the RDN; it cannot change other parts of a DN. For example, the
entry cn=Sue Jacobs,ou=People,dc=exam ple,dc=com can be changed to cn=Susan
Jacobs,ou=People,dc=exam ple,dc=com , but it cannot be modified to be cn=Sue
Jacobs,ou=old em ployees,dc=exam ple,dc=com .
T he following command renames Sue Jacobs to Susan Jacobs:
dn: cn=Sue Jacobs,ou=Marketing,dc=example,dc=com
changetype: modrdn
newrdn: cn=Susan Jacobs
deleteoldrdn: 0
Because deleteoldrdn is 0, this example retains the existing RDN as a value in the new entry. T he
resulting entry would therefore have a common name (cn) attribute set to both Sue Jacobs and Susan
Jacobs, in addition to all the other attributes included in the original entry. However, using the following
command causes the server to delete cn=Sue Jacobs, so that only cn=Susan Jacobs remains in
the entry:
dn: cn=Sue Jacobs,ou=Marketing,dc=example,dc=com
changetype: modrdn
newrdn: cn=Susan Jacobs
deleteoldrdn: 1
3.3.2.1. A Note on Renaming Entries
T he m odrdn change type cannot move an entry to a completely different subtree. T o move an entry to a
completely different branch, you must create a new entry in the alternative subtree using the old entry's
attributes, and then delete the old entry.
Also, for the same reasons that you cannot delete an entry if it is a branch point, you cannot rename an
entry if it has any children. Doing so would orphan the children in the tree, which is not allowed by the
LDAP protocol. For example, of the following three entries, only the last two entries can be renamed:
ou=People,dc=example,dc=com
cn=Paula Simon,ou=People,dc=example,dc=com
cn=Jerry O'Connor,ou=People,dc=example,dc=com
125
Red Hat D irectory Server 8.2 Administration Guide
T he entry that identifies the People subtree can be renamed only if no other entries exist below it.
3.3.3. Modifying an Entry Using LDIF
changetype: m odify can add, replace, or remove attributes or attribute values in an entry. When you
specify changetype: m odify, you must also provide a change operation to indicate how the entry is
to be modified. Change operations can be as follows:
add: attribute
Adds the specified attribute or attribute value. If the attribute type does not currently exist for the
entry, then the attribute and its corresponding value are created. If the attribute type already exists
for the entry, then the specified attribute value is added to the existing value. If the particular attribute
value already exists for the entry, then the operation fails, and the server returns an error.
replace: attribute
T he specified values are used to entirely replace the attribute's values. If the attribute does not
already exist, it is created. If no replacement value is specified for the attribute, the attribute is
deleted.
delete: attribute
T he specified attribute is deleted. If more than one value of an attribute exists for the entry, then all
values of the attribute are deleted in the entry. T o delete just one of many attribute values, specify
the attribute and associated value on the line following the delete change operation.
T his section contains the following topics:
Section 3.3.3.1, “Adding Attributes to Existing Entries Using LDIF”
Section 3.3.3.2, “Changing an Attribute Value Using LDIF”
Section 3.3.3.3, “Deleting All Values of an Attribute Using LDIF”
Section 3.3.3.4, “Deleting a Specific Attribute Value Using LDIF”
3.3.3.1. Adding Attributes to Existing Entries Using LDIF
Using changetype: m odify with the add operation cam add an attribute and an attribute value to an
entry. For example, the following LDIF update statement adds a telephone number to the entry:
dn: cn=Barney Fife,ou=People,dc=example,dc=com
changetype: modify
add: telephonenumber
telephonenumber: 555-1212
T he following example adds two telephone numbers to the entry:
dn: cn=Barney Fife,ou=People,dc=example,dc=com
changetype: modify
add: telephonenumber
telephonenumber: 555-1212
telephonenumber: 555-6789
T he following example adds two telephonenum ber attributes and a m anager attribute to the entry:
126
Chapter 3. Creating D irectory Entries
dn: cn=Barney Fife,ou=People,dc=example,dc=com
changetype: modify
add: telephonenumber
telephonenumber: 555-1212
telephonenumber: 555-6789
add: manager
manager: cn=Sally Nixon,ou=People,dc=example,dc=com
T he following example adds a jpeg photograph to the directory. In order to add this attribute to the
directory, use the -b parameter, which indicates that ldapm odify should read the referenced file for
binary values if the attribute value begins with a slash:
dn: cn=Barney Fife,ou=People,dc=example,dc=com
changetype: modify
add: jpegphoto
jpegphoto: /path/to/photo
You can also add a jpeg photograph to the directory using the following standard LDIF notation:
jpegphoto: < file:/path/to/photo
Using the standard notation means that the -b parameter does not need to be used withldapm odify.
However, you must add version:1 to the beginning of the LDIF file or with LDIF update statements. For
example:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
version: 1
dn: cn=Barney Fife,ou=People,dc=example,dc=com
changetype: modify
add: userCertificate
userCertificate;binary:< file: BarneysCert
NOTE
Standard LDIF notation can only be used with the ldapm odify command, not with other
command-line utilities.
3.3.3.2. Changing an Attribute Value Using LDIF
changetype: m odify with the replace operation changes all values of an attribute in an entry. For
example, the following LDIF update statement changes Barney's manager from Sally Nixon to Wally
Hensford:
dn: cn=Barney Fife,ou=People,dc=example,dc=com
changetype: modify
replace: manager
manager: cn=Wally Hensford,ou=People,dc=example,dc=com
If the entry has multiple instances of the attribute, then to change one of the attribute values, you must
delete the attribute value first and then add the replacement value. For example, this entry has two
telephone numbers:
127
Red Hat D irectory Server 8.2 Administration Guide
cn=Barney Fife,ou=People,dc=example,dc=com
objectClass: inetOrgPerson
cn: Barney Fife
sn: Fife
telephonenumber: 555-1212
telephonenumber: 555-6789
T o change the telephone number 555-1212 to 555-4 321, use the following LDIF update statement:
dn: cn=Barney Fife,ou=People,dc=example,dc=com
changetype: modify
delete: telephonenumber
telephonenumber: 555-1212
add: telephonenumber
telephonenumber: 555-4321
T he entry is now as follows:
cn=Barney Fife,ou=People,dc=example,dc=com
objectClass: inetOrgPerson
cn: Barney Fife
sn: Fife
telephonenumber: 555-6789
telephonenumber: 555-4321
3.3.3.3. Deleting All Values of an Attribute Using LDIF
changetype: m odify with the delete operation deletes an attribute from an entry. If the entry has
more than one instance of the attribute, you must indicate which of the attributes to delete.
For example, the following LDIF update statement deletes all instances of the telephonenumber
attribute from the entry, regardless of how many times it appears in the entry:
dn: cn=Barney Fife,ou=People,dc=example,dc=com
changetype: modify
delete: telephonenumber
T o delete just a specific instance of the telephonenumber attribute, simply delete that specific attribute
value, as described in the next section.
3.3.3.4 . Deleting a Specific Attribute Value Using LDIF
Running changetype: m odify with the delete operation can delete a single value for an attribute
value from an entry, as well as deleting all instances of the attribute. For example, consider the following
entry:
cn=Barney Fife,ou=People,dc=example,dc=com
objectClass: inetOrgPerson
cn: Barney Fife
sn: Fife
telephonenumber: 555-1212
telephonenumber: 555-6789
T o delete the 555-1212 telephone number from this entry, use the following LDIF update statement:
128
Chapter 3. Creating D irectory Entries
dn: cn=Barney Fife,ou=People,dc=example,dc=com
changetype: modify
delete: telephonenumber
telephonenumber: 555-1212
Barney's entry then becomes:
cn=Barney Fife,ou=People,dc=example,dc=com
objectClass: inetOrgPerson
cn: Barney Fife
sn: Fife
telephonenumber: 555-6789
3.3.4. Deleting an Entry Using LDIF
changetype: delete is the change type which deletes an entire entry from the directory.
NOTE
You can only delete leaf entries. T herefore, when you delete an entry, make sure that no other
entries exist under that entry in the directory tree. T hat is, you cannot delete an organizational
unit entry unless you have first deleted all the entries that belong to the organizational unit.
For example, of the following three entries, only the last two entries can be deleted:
ou=People,dc=example,dc=com
cn=Paula Simon,ou=People,dc=example,dc=com
cn=Jerry O'Connor,ou=People,dc=example,dc=com
T he entry that identifies the People subtree can be deleted only if no other entries exist below it.
T he following LDIF update statements can be used to delete person entries:
dn: cn=Pete Minsky,ou=People,dc=example,dc=com
changetype: delete
dn: cn=Sue Jacobs,ou=People,dc=example,dc=com
changetype: delete
WARNING
Do not delete the suffix o=NetscapeRoot. T he Admin Server uses this suffix to store
information about installed Directory Servers. Deleting this suffix could force you to reinstall the
Directory Server.
3.3.5. Modifying an Entry in an Internationalized Directory
If the attribute values in the directory are associated with languages other than English, the attribute
values are associated with language tags. When using the ldapm odify command-line utility to modify
an attribute that has an associated language tag, you must match the value and language tag exactly or
the modify operation will fail.
129
Red Hat D irectory Server 8.2 Administration Guide
For example, to modify an attribute value that has a language tag of lang-fr, include lang-fr in the
modify operation, as follows:
dn: bjensen,dc=example,dc=com
changetype: modify
replace: homePostalAddress;lang-fr
homePostalAddress;lang-fr: 34 rue de Seine
3.4. Tracking Modifications to Directory Entries
It can be useful to track when changes are made to entries. T here are two aspects of entry modifications
that the Directory Server tracks:
Using change sequence numbers to track changes to the database. T his is similar to change
sequence numbers used to in replication and synchronization. Every normal directory operation
triggers a sequence number.
Assigning creation and modification information. T hese attributes record the names of the user who
created and most recently modified an entry, as well as the timestamps of when it was created and
modified.
NOTE
T he entry USN, modify time and name, and create time and name are all operational attributes
and are not returned in a regular ldapsearch. For details on running a search for operational
attributes, see Section 8.4.7, “Searching for Operational Attributes”.
3.4.1. Tracking Modifications to the Database through Update Sequence
Numbers
T he Entry USN Plug-in provides a way for LDAP clients to know that something — anything — in the
database has changed.
When the plug-in is enabled, an update sequence number (USN) is assigned to an entry every time a
write operation occurs. (Write operations include add, modify, modrdn, and delete operations. Internal
database operations, like export operations, are not counted in the update sequence.) T he USN is
evaluated globally, for the entire database, not for the single entry.
NOTE
T he USN counter is local to each backend and there is only a single counter per instance. T his
means that there is only a single local instance of the Entry USN plug-in.
T he entry shows the change number for the last modification to that entry in the entryUSN attribute. T he
root DSE contains another attribute which contains the most recent USN assigned to any entry in the
database, in the lastusn attribute.
130
Chapter 3. Creating D irectory Entries
Example 3.1. Example Entry USN
dn: uid=jsmith,ou=People,dc=example,dc=com
mail: [email protected]
uid: jsmith
givenName: John
objectClass: top
objectClass: person
objectClass: organizationalPerson
objectClass: inetorgperson
sn: Smith
cn: John Smith
userPassword: {SSHA}EfhKCI4iKl/ipZMsWlITQatz7v2lUnptxwZ/pw==
entryusn: 1122
T he USN is similar to the change sequence number for replication and synchronization, in that it simply
ticks upward to track changes. T he entry USN is maintained separately from the CSNs, however.
T he Entry USN Plug-in does have some implications for replication.
Replication operations are treated by the Entry USN Plug-in as regular modify operations. T his means
that the replication process itself triggers the entryUSN to increment.
For single-master replication, if the entryUSN attribute is replicated, the consumer has the same
entryUSNs as the supplier. T here is no conflict, because the consumer server must be read-only, so
there is no point in having a local entry USN counter. When the consumer is initialized, it is initialized with
all of the entry USN numbers from the supplier (unless entryUSN is excluded using fractional
replication) and continues from there.
For multi-master replication, however, it is possible for there to be different USNs on entries across both
suppliers and consumers. If the entry USN is replicated, it is simply overwritten by the local Entry USN
Plug-in or locally ignored, if the plug-in is not enabled.
T he simplest solution is to exclude the entryUSN attribute using fractional replication (see Section 9.1.7,
“Replicating a Subset of Attributes with Fractional Replication”), so that all entryUSN numbers are local.
When the entryUSN attribute is excluded in the replication agreement, then when the consumers are
initialized, they have entryUSNs that start at zero (0) locally instead of using the supplier's entryUSNs.
3.4 .1.1. Configuring the USN Plug-in
T he Entry USN Plug-in must be enabled for USNs to be recorded on entries, as described in
Section 1.10.1, “Enabling Plug-ins”. T he plug-in can be enabled through the Directory Server Console or
through the command line. For example:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389
dn: cn=USN,cn=plugins,cn=config
changetype: modify
replace: nsslapd-pluginEnabled
nsslapd-pluginEnabled: on
T hen restart the server to apply the changes.
3.4 .1.2. Cleaning up USN T ombstone Entries
131
Red Hat D irectory Server 8.2 Administration Guide
T he Entry USN Plug-in moves entries to tombstone entries when the entry is deleted. If replication is
enabled, then separate tombstone entries are kept by both the Entry USN and Replication Plug-ins. Both
tombstone entries are deleted by the replication process, but for server performance, it can be beneficial
to delete the entry USN tombstones before converting a server to a replica or to free memory for the
server.
T he usn-tom bstone-cleanup.pl command deletes USN tombstone entries for a specific database
backend or specific suffix. Optionally, it can delete all of tombstone entries up to a certain USN. For
example:
/usr/lib64/dirsrv/instance_name/usn-tombstone-cleanup.pl -D "cn=directory
manager" -w secret -s "ou=people,dc=example,dc=com" -m 1100
Either the backend must be specified using the -n option or the suffix, using the -s option. If both are
given, then the suffix in the -s option is used.
T he options for usn-tom bstone-cleanup.pl command are listed in T able 3.5, “usn-tombstonecleanup.pl Options”. More details for this tool are in the Configuration and Command-Line Tool
Reference.
T able 3.5. usn-tombstone-cleanup.pl Options
Option
Description
-D rootdn
Gives the user DN with root permissions, such
as Directory Manager. T he default is the DN of
the Directory Manager, which is read from the
nsslapd-root attribute under cn=config.
-m maximum_USN
Sets the upper bound for entries to delete. All
tombstone entries with an entryUSN value up to
the specified maximum (inclusive) are deleted, but
not past that USN value. If no maximum USN
value is set, then all backend tombstone entries
are deleted.
-n backendInstance
Gives the name of the database containing the
entries to clean (delete).
-s suffix
Gives the name of the suffix containing the
entries to clean (delete).
-w password
T he password associated with the user DN.
3.4.2. Tracking Entry Modifications through Operational Attributes
T he Directory Server can maintain some special operational attributes for every directory entry, showing
basic creation and modification information:
creatorsName. T he distinguished name of the person who initially created the entry.
createTimestamp. T he timestamp for when the entry was created in GMT (Greenwich Mean T ime)
format.
modifiersName. T he distinguished name of the person who last modified the entry.
modifyTimestamp. T he timestamp for when the entry was last modified in GMT format.
Operational attributes are not returned in default directory searches and must be explicitly requested, as
described in Section 8.4.7, “Searching for Operational Attributes”.
132
Chapter 3. Creating D irectory Entries
described in Section 8.4.7, “Searching for Operational Attributes”.
NOTE
When a database link is used by a client application to create or modify entries, the
creatorsName and modifiersName attributes do not reflect the real creator or modifier of the
entries. T hese attributes contain the name of the administrator who is granted proxy
authorization rights on the remote server. For information on proxy authorization, see
Section 2.3.1.2.2, “Providing Bind Credentials”.
T o enable the Directory Server to track when entries are created or modified:
1. In the Directory Server Console, select the Configuration tab, and then select the top entry in
the navigation tree in the left pane.
2. Select the Settings tab in the right pane.
3. Select the T rack Entry Modification T im es checkbox.
T he server adds the creatorsName, createTimestamp, modifiersName, and modifyTimestamp
attributes to every newly created or modified entry.
4. Open the T asks tab, and click Restart Directory Server.
133
Red Hat D irectory Server 8.2 Administration Guide
NOTE
T he Directory Server must be restarted for the changes to take effect.
3.5. Maintaining Referential Integrity
Referential integrity is a database mechanism that ensures relationships between related entries are
maintained. In the Directory Server, the referential integrity can be used to ensure that an update to one
entry in the directory is correctly reflected in any other entries that may refer to the updated entry.
For example, if a user's entry is removed from the directory and referential integrity is enabled, the
server also removes the user from any groups of which the user is a member. If referential integrity is
not enabled, the user remains a member of the group until manually removed by the administrator. T his
is an important feature if you are integrating the Directory Server with other products that rely on the
directory for user and group management.
3.5.1. How Referential Integrity Works
When the Referential Integrity Plug-in is enabled, it performs integrity updates on specified attributes
immediately after a delete or rename operation. By default, the Referential Integrity Plug-in is disabled.
NOTE
T he Referential Integrity Plug-in should only be enabled on one supplier replica in a multi-master
replication environment to avoid conflict resolution loops. When enabling the plug-in on servers
issuing chaining requests, be sure to analyze performance resource and time needs, as well as
your integrity needs. Integrity checks can be time-consuming and draining on memory and CPU.
Whenever a user or group entry is deleted or renamed in the directory, the operation is logged to the
referential integrity log file (/var/log/dirsrv/slapd-instance_name). After a specified time, known
as the update interval, the server performs a search on all attributes for which referential integrity is
enabled and matches the entries resulting from that search with the DNs of deleted or modified entries
present in the log file. If the log file shows that the entry was deleted, the corresponding attribute is
deleted. If the log file shows that the entry was changed, the corresponding attribute value is modified
134
Chapter 3. Creating D irectory Entries
accordingly.
By default, when the Referential Integrity Plug-in is enabled, it performs integrity updates on the member,
uniquemember, owner, and seeAlso attributes immediately after a delete or rename operation. However,
the behavior of the Referential Integrity Plug-in can be configured to suit the needs of the directory in
several different ways:
Record referential integrity updates in the replication changelog.
Modify the update interval.
Select the attributes to which to apply referential integrity.
Disable referential integrity.
All attributes used in referential integrity must be indexed for presence and equality; not indexing those
attributes results poor server performance for modify and delete operations. See Section 7.2, “Creating
Standard Indexes” for more information about checking and creating indexes.
3.5.2. Using Referential Integrity with Replication
T here are certain limitations when using the Referential Integrity Plug-in in a replication environment:
Never enable it on a dedicated consumer server (a server that contains only read-only replicas).
Never enable it on a server that contains a combination of read-write and read-only replicas.
It is possible to enable it on a supplier server that contains only read-write replicas.
With multi-master replication, enable the plug-in on just one supplier.
If the replication environment satisfies the all of those condition, you can enable the Referential Integrity
Plug-in.
1. Enable the Referential Integrity Plug-in as described in Section 3.5.3, “Enabling and Disabling
Referential Integrity”.
2. Configure the plug-in to record any integrity updates in the changelog.
3. Ensure that the Referential Integrity Plug-in is disabled on all consumer servers.
NOTE
Because the supplier server sends any changes made by the Referential Integrity Plug-in
to consumer servers, it is unnecessary to run the Referential Integrity Plug-in on consumer
servers.
3.5.3. Enabling and Disabling Referential Integrity
3.5.3.1. Enabling and Disabling Referential Integrity in the Console
1. Select the Configuration tab, and expand the Plugins folder.
2. Select Referential Integrity Postoperation Plug-in from the list.
135
Red Hat D irectory Server 8.2 Administration Guide
3. Check the Enable plugin checkbox to enable the plug-in; clear it to disable it.
4. Fill in the correct path to the plug-in by default; plug-ins are located in
/usr/lib64 /dirsrv/plugins.
5. Restart the Directory Server to apply the changes. In the T asks tab, select Restart the
Directory Server.
136
Chapter 3. Creating D irectory Entries
3.5.3.2. Enabling and Disabling Referential Integrity from the Command Line
T o disable or enable the Referential Integrity Plug-in:
1. Use ldapm odify to edit the value of the nsslapd-pluginEnabled attribute. For example:
[5]
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=referential integrity postoperation,cn=plugins,cn=config
changetype: modify
replace: nsslapd-pluginEnabled
nsslapd-pluginEnabled: on
2. T hen restart the server.
service dirsrv restart
3.5.4. Modifying the Update Interval
By default, the server makes referential integrity updates immediately after a delete or a m odrdn
operation. T o reduce the impact this operation has on your system, increase the amount of time between
updates. Although there is no maximum update interval, the following intervals are commonly used:
Update immediately
90 seconds
3600 seconds (updates occur every hour)
10,800 seconds (updates occur every 3 hours)
28,800 seconds (updates occur every 8 hours)
86,400 seconds (updates occur once a day)
604,800 seconds (updates occur once a week)
3.5.4 .1. Modifying the Update Interval from the Console
1. Select the Configuration tab, and expand the Plugins folder. Select the Referential
Integrity Postoperation Plug-in.
137
Red Hat D irectory Server 8.2 Administration Guide
2. In the arguments list, replace the value in the first text box with the appropriate time interval.
3. Restart the Directory Server to apply the changes. In the T asks tab, select Restart the
Directory Server.
3.5.4 .2. Modifying the Update Interval from the Command Line
1. Use ldapm odify[5] to edit the value of the nsslapd-pluginarg attribute. For example:
T he first argument listed sets the update intervale for referntial integrity checks. T o change the
interval, replace the nsslapd-pluginarg0 attribute.
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389
dn: cn=referential integrity postoperation,cn=plugins,cn=config
changetype: modify
replace: nsslapd-pluginarg0
nsslapd-pluginarg0: 600
2. T hen restart the server.
service dirsrv restart
3.5.5. Modifying the Attribute List
3.5.5.1. Modifying the Attribute List from the Console
By default, the Referential Integrity Plug-in is set up to check for and update the member, uniquemember,
owner, and seeAlso attributes. You can add or delete attributes to be updated through the Directory
Server Console, such as adding the nsroledn attribute if roles are being used.
NOTE
Keep in mind that any attribute specified in the Referential Integrity Plug-in parameter list must
have equality indexing on all databases. Otherwise, the plug-in scans every entry of the
databases for matching the deleted or modified DN, degrading performance severely. If you add
an attribute, ensure that it is indexed in all the backends.
NOTE
Improve the performance by removing any unused attributes from the list.
1. Select the Configuration tab, and expand the Plugins folder. Select the Referential
138
Chapter 3. Creating D irectory Entries
Integrity Postoperation Plug-in.
2. In the Argum ents section, use the Add and Delete buttons to modify the attributes in the list.
3. Restart the Directory Server to apply the changes. In the T asks tab, select Restart the
Directory Server.
NOTE
All attributes used in referential integrity must be indexed for presence and equality; not indexing
those attributes results poor server performance for modify and delete operations. See
Section 7.2, “Creating Standard Indexes” for more information about checking and creating
indexes.
3.5.5.2. Modifying the Attribute List from the Command Line
By default, the Referential Integrity Plug-in is set up to check for and update the member, uniquemember,
owner, and seeAlso attributes.
nsslapd-pluginarg3:
nsslapd-pluginarg4:
nsslapd-pluginarg5:
nsslapd-pluginarg6:
member
uniquemember
owner
seeAlso
T o add another attribute to the list:
1. Use ldapm odify[5] to add another attribute to the list by adding another nsslapd-pluginarg
attribute.
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389
dn: cn=referential integrity postoperation,cn=plugins,cn=config
changetype: modify
add: nsslapd-pluginarg7
nsslapd-pluginarg7: nsroledn
2. T hen restart the server.
service dirsrv restart
139
Red Hat D irectory Server 8.2 Administration Guide
NOTE
Keep in mind that any attribute specified in the Referential Integrity Plug-in parameter list must
have equality indexing on all databases. Otherwise, the plug-in scans every entry of the
databases for matching the deleted or modified DN, degrading performance severely. If you add
an attribute, ensure that it is indexed in all the backends.
[4] The LDAP to o ls referenc ed in this g uid e are Mo z illa LDAP, ins talled with Direc to ry Server in the /usr/l i b 6 4 /m o zl d ap
d irec to ry o n Red Hat Enterp ris e Linux 5 (6 4-b it); d irec to ries fo r o ther p latfo rms are lis ted in Sec tio n 1.3, “ LDAP To o l Lo c atio ns ” .
Ho wever, Red Hat Enterp ris e Linux s ys tems als o inc lud e LDAP to o ls fro m O p enLDAP. It is p o s s ib le to us e the O p enLDAP
c o mmand s as s ho wn in the examp les , b ut yo u mus t us e the - x arg ument to d is ab le SASL and allo w s imp le authentic atio n.
[5] The LDAP to o ls referenc ed in this g uid e are Mo z illa LDAP, ins talled with Direc to ry Server in the /usr/l i b 6 4 /m o zl d ap
d irec to ry o n Red Hat Enterp ris e Linux 5 (6 4-b it); d irec to ries fo r o ther p latfo rms are lis ted in Sec tio n 1.3, “ LDAP To o l Lo c atio ns ” .
Ho wever, Red Hat Enterp ris e Linux s ys tems als o inc lud e LDAP to o ls fro m O p enLDAP. It is p o s s ib le to us e the O p enLDAP
c o mmand s as s ho wn in the examp les , b ut yo u mus t us e the - x arg ument to d is ab le SASL and allo w s imp le authentic atio n.
14 0
Chapter 4. Populating D irectory D atabases
Chapter 4. Populating Directory Databases
Databases contain the directory data managed by the Red Hat Directory Server.
4.1. Importing Data
Directory Server can populate a database with data in one of two ways: by importing data (either through
the Directory Server Console or using the import tools) or by initializing a database for replication.
T able 4.1, “Import Method Comparison” describes the differences between an import and initializing
databases.
T able 4 .1. Import Method Comparison
Action
Import
Initialize Database
Overwrites database
No
Yes
LDAP operations
Add, modify, delete
Add only
Performance
More time-consuming
Fast
Partition specialty
Works on all partitions
Local partitions only
Response to server failure
Best effort (all changes made up
to the point of the failure remain)
Atomic (all changes are lost
after a failure)
LDIF file location
Local to Console
Local to Console or local to
server
Imports configuration
information (cn=config)
Yes
No
4.1.1. Importing Entries with Large Attributes
T he nsslapd-cachememsize attribute defines the size allowed for the entry cache.
T he import buffer is automatically set to 80% of the cache memory size setting. If the memory cache is
1GB, for example, then the import buffer is 800MB.
When importing a very large database or entries with large attributes (often with values like binary data
like certificate chains, CRLs, or images), then set the nsslapd-cachememsize attribute high enough so
that the import buffer has enough memory to process the entries.
4.1.2. Importing a Database from the Console
When performing an import operation from the Directory Server Console, an ldapm odify operation is
executed to append data, as well as to modify and delete entries. T he operation is performed on all of
the databases managed by the Directory Server and on remote databases to which the Directory Server
has a configured database link.
Import operations can be run on a server instance that is local to the Directory Server Console or on a
different host machine (a remote import operation).
You must be logged in as the Directory Manager in order to perform an import.
14 1
Red Hat D irectory Server 8.2 Administration Guide
NOTE
T he LDIF files used for import operations must use UT F-8 character set encoding. Import
operations do not convert data from local character set encoding to UT F-8 characterset encoding.
WARNING
All imported LDIF files must also contain the root suffix.
T o import data from the Directory Server Console:
1. Select the T asks tab. Scroll to the bottom of the screen, and select Im port Database.
Alternatively, open the Configuration tab and select Im port from the Console menu.
2. In the Im port Database dialog box, enter the full path to the LDIF file to import in the LDIF
file field, or click Browse to select the file to import.
If the Console is running on a machine remote to the directory, the field name appears as LDIF
file (on the m achine running the Console). When browsing for a file, you are not
browsing the current directory for the Directory Server host, but the filesystem of the machine
running the Console.
When importing a database through a remote Console, do not use a relative path to the database.
For remote imports, the operation fails with the error Cannot write to file... if a relative path is given
for the file. Always use an absolute path for remote import operations.
3. In the Options box, select one or both of the following options:
Add Only. T he LDIF file may contain modify and delete instructions in addition to the default
add instructions. For the server to ignore operations other than add, select the Add only
checkbox.
Continue on Error. Select the Continue on error checkbox for the server to continue with
the import even if errors occur. For example, use this option to import an LDIF file that contains
some entries that already exist in the database in addition to new ones. T he server notes
existing entries in the rejects file while adding all new entries.
4. In the File for Rejects field, enter the full path to the file in which the server is to record all
14 2
Chapter 4. Populating D irectory D atabases
entries it cannot import, or click Browse to select the file which will contain the rejects.
A reject is an entry which cannot be imported into the database; for example, the server cannot
import an entry that already exists in the database or an entry that has no parent object. T he
Console will write the error message sent by the server to the rejects file.
Leaving this field blank means the server will not record rejected entries.
T he server performs the import and also creates indexes.
NOTE
T railing spaces are dropped during a remote Console import but are preserved during both local
Console or ldif2db import operations.
4.1.3. Initializing a Database from the Console
T he existing data in a database can be overwritten by initializing databases.
You must be logged in as the Directory Manager in order to initialize a database because an LDIF
file that contains a root entry cannot be imported into a database except as the Directory Manager (root
DN). Only the Directory Manager has access to the root entry, such as dc=exam ple,dc=com .
WARNING
When initializing databases from an LDIF file, be careful not to overwrite the o=NetscapeRoot
suffix unless you are restoring data. Otherwise, initializing the database deletes information and
may require re-installing the Directory Server.
T o initialize a database using the Directory Server Console:
1. Select the Configuration tab.
2. Expand the Data tree in the left navigation pane. Expand the suffix of the database to initialize,
then click the database itself.
3. Right-click the database, and select Initialize Database.
14 3
Red Hat D irectory Server 8.2 Administration Guide
Alternatively, select Initialize Database from the Object menu.
4. In the LDIF file field, enter the full path to the LDIF file to import, or click Browse.
5. If the Console is running from a machine local to the file being imported, click OK and proceed with
the import immediately. If the Console is running from a machine remote to the server containing
the LDIF file, select one of the following options, then click OK:
From local machine. Indicates that the LDIF file is located on the local machine.
From server machine. Indicates that the LDIF file is located on a remote server.
T he default LDIF directory is /var/lib/dirsrv/slapd-instance_name/ldif.
4.1.4. Importing from the Command Line
T here are four methods for importing data through the command line:
Using ldif2db. T his import method overwrites the contents of the database and requires the server to
be stopped; see Section 4.1.4.1, “Importing Using the ldif2db Command-Line Script”.
Using ldif2db.pl. T his import method overwrites the contents of the database while the server is still
running; see Section 4.1.4.2, “Importing Using the ldif2db.pl Perl Script”.
Using ldif2ldap. T his method appends the LDIF file through LDAP. T his method is useful to append
data to all of the databases; see Section 4.1.4.3, “Importing Using the ldif2ldap Command-Line Script”.
Creating a cn=tasks entry. T his method creates a temporary task entry which automatically launches
an import operation. T his is functionally like running ldif2db. See Section 4.1.4.4, “Importing
through the cn=tasks Entry”.
NOTE
T he LDIF files used for import operations must use UT F-8 character set encoding. Import
operations do not convert data from local character set encoding to UT F-8 characterset encoding.
WARNING
All imported LDIF files must also contain the root suffix.
NOTE
T o import a database that has been encrypted, use the -E option with the script. See
Section 2.2.3.5, “Exporting and Importing an Encrypted Database” for more information.
4 .1.4 .1. Importing Using the ldif2db Command-Line Script
T he ldif2db script overwrites the data in the specified database. Also, the script requires that the
14 4
Chapter 4. Populating D irectory D atabases
Directory Server be stopped when the import begins.
By default, the script first saves and then merges any existing o=NetscapeRoot configuration
information with the o=NetscapeRoot configuration information in the files being imported.
WARNING
T his script overwrites the data in the database.
T o import LDIF:
1. Stop the server.
service dirsrv stop instance
2. Open the Directory Server instance directory.
cd /etc/dirsrv/slapd-instance_name
3. Run the ldif2db command-line script.
ldif2db -n Database1 -i /var/lib/dirsrv/slapd-instance_name/ldif/demo.ldif
-i /var/lib/dirsrv/slapd-instance_name/ldif/demo2.ldif
For more information about using this script, see the Directory Server Configuration and
Command-Line Tool Reference.
WARNING
If the database specified in the -n option does not correspond with the suffix contained by
the LDIF file, all of the data contained by the database is deleted, and the import fails. Make
sure that the database name is not misspelled.
T able 4 .2. ldif2db Parameters
Option
Description
-i
Specifies the full path name of the LDIF files to be
imported. T his option is required. T o import more
than one LDIF file at a time, use multiple -i
arguments. When multiple files are imported, the
server imports the LDIF files in the order which
they are specified from the command line.
-n
Specifies the name of the database to which to
import the data.
For more information about using this script, see the Directory Server Configuration and Command-Line
Tool Reference.
4 .1.4 .2. Importing Using the ldif2db.pl Perl Script
As with the ldif2db script, the ldif2db.pl script overwrites the data in the specified database. T his
14 5
Red Hat D irectory Server 8.2 Administration Guide
script requires the server to be running in order to perform the import.
WARNING
T his script overwrites the data in the database.
1. Open the Directory Server instance directory.
cd /etc/dirsrv/slapd-instance_name
2. Run the ldif2db script.
ldif2db -D "cn=Directory Manager" -w secret -i
/var/lib/dirsrv/slapd-instance_name/ldif/demo.ldif -n Database1
For more information about using this script, see the Directory Server Configuration and
Command-Line Tool Reference.
NOTE
You do not need root privileges to run the script, but you must authenticate as the
Directory Manager.
T able 4 .3. ldif2db Options
Option
Description
-D
Specifies the DN of the administrative user.
-w
Specifies the password of the administrative user.
-i
Specifies the LDIF files to be imported. T his
option is required. T o important multiple LDIF files
at a time, use multiple -i arguments. When
multiple files are imported, the server imports the
LDIF files in the order they are specified in the
command line.
-n
Specifies the name of the database to which to
import the data.
4 .1.4 .3. Importing Using the ldif2ldap Command-Line Script
T he ldif2ldap script appends the LDIF file through LDAP. Using this script, data are imported to all
directory databases at the same time. T he server must be running in order to import using ldif2ldap.
T o import LDIF using ldif2ldap:
1. Open the Directory Server instance directory:
cd /etc/dirsrv/slapd-instance_name
2. Run the ldif2ldap command-line script.
14 6
Chapter 4. Populating D irectory D atabases
ldif2ldap "cn=Directory Manager" secretpwd
/var/lib/dirsrv/slapd-instance_name/ldif/demo.ldif
T he ldif2ldap script requires the DN of the administrative user, the password of the
administrative user, and the absolute path and filename of the LDIF files to be imported.
For more information about using this script, see the Directory Server Configuration and
Command-Line Tool Reference.
4 .1.4 .4 . Importing through the cn=tasks Entry
T he cn=tasks,cn=config entry in the Directory Server configuration is a container entry for
temporary entries that the server uses to manage tasks. Several common directory tasks have container
entries under cn=tasks,cn=config. T emporary task entries can be created under
cn=im port,cn=tasks,cn=config to initiate an import operation.
As with the ldif2db and ldif2db.pl scripts, an import operation in cn=tasks overwrites all of the
information in the database.
T his task entry requires three attributes:
A unique name (cn)
T he filename of the LDIF file to import (nsFilename)
T he name of the database into which to import the file (nsInstance)
It is also possible to supply the DNs of suffixes to include or exclude from the import, analogous to the s and -x options, respectively, for the ldif2db and ldif2db.pl scripts.
T he entry is simply added using ldapm odify, as described in Section 3.2.4, “Adding and Modifying
Entries Using ldapmodify”. For example:
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=example import,cn=import,cn=tasks,cn=config
objectclass: extensibleObject
cn: example import
nsFilename: /home/files/example.ldif
nsInstance: userRoot
nsIncludeSuffix: ou=People,dc=example,dc=com
nsExcludeSuffix: ou=Groups,dc=example,dc=com
As soon as the task is completed, the entry is removed from the directory configuration.
T he Directory Server Configuration and Command-Line Tool Reference has more information on the
available attributes for running Directory Server import tasks under the cn=tasks entries.
4.2. Exporting Data
LDAP Data Interchange Format (LDIF) files are used to export database entries from the Directory Server
databases. LDIF is a standard format described in RFC 2849, The LDAP Data Interchange Format
(LDIF) - Technical Specification.
Exporting data can be useful for the following:
Backing up the data in the database.
14 7
Red Hat D irectory Server 8.2 Administration Guide
Copying data to another Directory Server.
Exporting data to another application.
Repopulating databases after a change to the directory topology.
For example, if a directory contains one database, and its contents are split into two databases, then the
two new databases receive their data by exporting the contents of the old databases and importing it
into the two new databases, as illustrated in Figure 4.1, “Splitting a Database Contents into T wo
Databases”.
NOTE
T he export operations do not export the configuration information (cn=config), schema
information (cn=schem a), or monitoring information (cn=m onitor).
Figure 4 .1. Splitting a Database Contents into T wo Databases
T he Directory Server Console or command-line utilities can be used to export data.
Section 4.2.1, “Exporting Directory Data to LDIF Using the Console”
Section 4.2.2, “Exporting a Single Database to LDIF Using the Console”
Section 4.2.3, “Exporting to LDIF from the Command Line”
WARNING
Do not stop the server during an export operation.
4.2.1. Exporting Directory Data to LDIF Using the Console
Some or all of directory data can be exported to LDIF, depending upon the location of the final exported
14 8
Chapter 4. Populating D irectory D atabases
file. When the LDIF file is on the server, only the data contained by the databases local to the server can
be exported. If the LDIF file is remote to the server, all of the databases and database links can be
exported.
Export operations can be run to get data from a server instance that is local to the Directory Server
Console or from a different host machine (a remote export operation).
Export directory data to LDIF from the Directory Server Console while the server is running:
1. Select the T asks tab. Scroll to the bottom of the screen, and click Export Database(s).
Alternatively, select the Configuration tab and click the Export from the Console menu.
2. Enter the full path and filename of the LDIF file in the LDIF File field, or click Browse to locate
the file.
Browse is not enabled if the Console is running on a remote server. When the Browse button is
not enabled, the file is stored in the default directory,
/var/lib/dirsrv/slapd-instance_name/ldif.
3. If the Console is running on a machine remote to the server, two radio buttons are displayed
beneath the LDIF File field.
Select T o local m achine to export the data to an LDIF file on the machine from which the
Console is running.
Select T o server m achine to export to an LDIF file located on the server's machine.
4. T o export the whole directory, select the Entire database radio button.
T o export only a single subtree of the suffix contained by the database, select the Subtree radio
button, and then enter the name of the suffix in the Subtree text box. T his option exports a
subtree that is contained by more than one database.
Alternatively, click Browse to select a suffix or subtree.
4.2.2. Exporting a Single Database to LDIF Using the Console
It is also possible to export a single database to LDIF. Do the following while the server is running:
1. Select the Configuration tab.
2. Expand the Data tree in the left navigation pane. Expand the suffix, and select the database
under the suffix.
3. Right-click the database, and select Export Database.
14 9
Red Hat D irectory Server 8.2 Administration Guide
Alternatively, select Export Database from the Object menu.
4. In the LDIF file field, enter the full path to the LDIF file, or click Browse.
When the Browse button is not enabled, the file is stored in the default directory,
/var/lib/dirsrv/slapd-instance_name/ldif.
4.2.3. Exporting to LDIF from the Command Line
T here are three methods for exporting data through the command line:
Using db2ldif. T his method runs the command-line utility; unlike the import script, ldif2db, this utility
can be run while the server is running.
Using db2ldif.pl. T his Perl script behaves the same as the db2ldif command-line utility and takes
the same options.
Creating a cn=tasks entry. T his method creates a temporary task entry which automatically launches
an export operation. T his is functionally like running db2ldif, with one exception: when running
db2ldif or db2ldif.pl for a replica (with a -r option, the server must be stopped first. T he
cn=tasks entry can be added and export replica information while the server is still running. See
Section 4.2.3.2, “Exporting through the cn=tasks Entry”.
4 .2.3.1. Exporting a Database Using db2ldif or db2ldif.pl
Databases can be exported to LDIF using the db2ldif command-line script or the db2ldif.pl Perl
script. Both of these tools export all of the database contents or a part of their contents to LDIF when the
server is running or stopped.
T hese script take the same options.
150
Chapter 4. Populating D irectory D atabases
NOTE
T he -E option is required to export a database that has been encrypted. See Section 2.2.3.5,
“Exporting and Importing an Encrypted Database” for more information.
NOTE
If the database being exported is a replica, then the server must be stopped before the export
script is run and the export script must have the -r.
T o export to LDIF from the command line:
1. Open the Directory Server instance directory:
cd /etc/dirsrv/slapd-instance_name
2. Run either the db2ldif command-line script or the db2ldif.pl Perl script. For example:
db2ldif -n database1 -a /export/output.ldif
T his exports the database contents to /export/output.ldif. If the -a option is not specified,
then the database information is exported to
/var/lib/dirsrv/slapd-instance_name/ldif/instance_name-database1-date.ldif.
For example:
db2ldif -n database1
It is also possible to specify which suffixes to export, using the -s option. For example:
db2ldif -s "dc=example,dc=com"
T he LDIF file in this case would be
/var/lib/dirsrv/slapd-instance_name/ldif/instance_name-exam ple2010_04 _30_112718.ldif, using the name of the suffix rather than the database.
If the suffix specified is a root suffix, such as dc=exam ple,dc=com , then it is not necessary to
specify the database or to use the -n option.
For more information about using these scripts, see the Directory Server Configuration and CommandLine Tool Reference.
151
Red Hat D irectory Server 8.2 Administration Guide
T able 4 .4 . db2ldif Options
Option
Description
-n
Specifies the name of the database from which
the file is being exported.
-s
Specifies the suffix or suffixes to include in the
export. If the suffix is a root suffix, such as
dc=exam ple,dc=com , then the -n option is not
required. T here can be multiple -s arguments.
-a
Defines the output file to which Directory Server
exports the LDIF. T his file must be an absolute
path. If the -a option is not given, the output ldif is
stored in the
/var/lib/dirsrv/slapd-instance_name/l
dif directory and is automatically named
serverIDdatabase-YYYY_MM_DD_hhmmxx.ldif with
the -n option or serverID-firstsuffixvalueYYYY_MM_DD_hhmmxx.ldif with the -s
option.
-r
Specifies that the exported database is a
consumer replica. In this case, the appropriate
settings and entries are included with the LDIF to
initialize the replica when the LDIF is imported.
-E
Decrypts an encrypted database so it can be
exported.
4 .2.3.2. Exporting through the cn=tasks Entry
T he cn=tasks,cn=config entry in the Directory Server configuration is a container entry for
temporary entries that the server uses to manage tasks. Several common directory tasks have container
entries under cn=tasks,cn=config. T emporary task entries can be created under
cn=export,cn=tasks,cn=config to initiate an export operation.
T he export task entry requires three attributes:
A unique name (cn)
T he filename of the LDIF file to which to export the database (nsFilename)
T he name of the database to export (nsInstance)
It is also possible to supply the DNs of suffixes to include or exclude from the export operation,
analogous to the -s and -x options, respectively, for the db2ldif and db2ldif.pl scripts.
Additionally, if the database is a replica, then the appropriate replica information can be included to
initialize the new consumer when the LDIF is imported; this is set in the nsExportReplica,
corresponding to the -r option.
T he entry is simply added using ldapm odify, as described in Section 3.2.4, “Adding and Modifying
Entries Using ldapmodify”. For example:
152
Chapter 4. Populating D irectory D atabases
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=example export,cn=export,cn=tasks,cn=config
objectclass: extensibleObject
cn: example export
nsInstance: userRoot
nsFilename: /home/files/example.ldif
nsExportReplica: true
nsIncludeSuffix: ou=People,dc=example,dc=com
nsExcludeSuffix: ou=Groups,dc=example,dc=com
As soon as the task is completed, the entry is removed from the directory configuration.
T he Directory Server Configuration and Command-Line Tool Reference has more information on the
available attributes for running Directory Server export tasks under the cn=tasks entries.
4.3. Backing up and Restoring Data
Databases can be backed up and restored using the Directory Server Console or a command-line script.
Section 4.3.1, “Backing up All Databases”
Section 4.3.2, “Backing up the dse.ldif Configuration File”
Section 4.3.3, “Restoring All Databases”
Section 4.3.4, “Restoring a Single Database”
Section 4.3.5, “Restoring Databases T hat Include Replicated Entries”
Section 4.3.6, “Restoring the dse.ldif Configuration File”
WARNING
Do not stop the server during a backup or restore operation.
4.3.1. Backing up All Databases
T he following procedures describe backing up all of the databases in the directory using the Directory
Server Console and from the command line.
NOTE
T hese backup methods cannot be used to back up the data contained by databases on a remote
server that are chained using database links.
4 .3.1.1. Backing up All Databases from the Console
When backing up databases from the Directory Server Console, the server copies all of the database
contents and associated index files to a backup location. A backup can be performed while the server is
running.
T o back up databases from the Directory Server Console:
1. Select the T asks tab.
153
Red Hat D irectory Server 8.2 Administration Guide
2. Click Back Up Directory Server.
3. Enter the full path of the directory to store the backup file in the Directory text box, or click Use
default, and the server provides a name for the backup directory.
If the Console is running on the same machine as the directory, click Browse to select a local
directory.
With the default location, the backup files are placed in
/var/lib/dirsrv/slapd-instance_name/bak. By default, the backup directory name
contains the name of the server instance and the time and date the backup was created
(instance_name-YYYY_MM_DD_hhmmss).
4 .3.1.2. Backing up All Databases from the Command Line
Databases can be backed up from the command line using either the db2bak command-line script or
the db2bak Perl script. T he Perl script works when the server is running or when the server is stopped;
the command-line script can only be run when the server is stopped.
Configuration information cannot be backed up using this backup method. For information on backing up
the configuration information, see Section 4.3.2, “Backing up the dse.ldif Configuration File”.
T o back up the directory from the command line using the db2bak script:
1. Stop the server instance.
cd service dirsrv stop slapd-example
TIP
T o avoid shutting down the server when running a backup, use the db2bak.pl Perl script
instead of the bd2bak tool. T hese are both located in the
/usr/lib[64 ]/dirsrv/slapd-exam ple directory.
2. Open the instance's tools directory. For example, on Red Hat Enterprise Linux 5 (64-bit):
cd /usr/lib64/dirsrv/slapd-instance
3. Run the db2bak command-line script, specifying the backup file name and directory.
154
Chapter 4. Populating D irectory D atabases
./db2bak /var/lib/dirsrv/slapd-instance_name/bak/instance_name2010_04_30_16_27_56
T he backup directory where the server saves the backed up databases can be specified with the
script. If a directory is not specified, the backup file is stored in
/var/lib/dirsrv/slapd-instance_name/bak. By default, the backup directory is named with
the Directory Server instance name and the date of the backup (serverIDYYYY_MM_DD_hhmmss).
4. Start the server instance again.
cd service dirsrv start slapd-example
For more information about using these scripts, see the Directory Server Configuration and CommandLine Tool Reference.
4 .3.1.3. Backing up the Database through the cn=tasks Entry
T he cn=tasks,cn=config entry in the Directory Server configuration is a container entry for
temporary entries that the server uses to manage tasks. Several common directory tasks have container
entries under cn=tasks,cn=config. T emporary task entries can be created under
cn=backup,cn=tasks,cn=config to initiate a backup operation.
T he backup task entry requires three attributes:
A unique name (cn).
T he directory to write the backup file to (nsArchiveDir). T he backup file is named with the Directory
Server instance name and the date of the backup (serverID-YYYY_MM_DD_hhmmss).
T he type of database (nsDatabaseType); the only option is ldbm database.
T he entry is simply added using ldapm odify, as described in Section 3.2.4, “Adding and Modifying
Entries Using ldapmodify”. For example:
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=example backup,cn=backup,cn=tasks,cn=config
objectclass: extensibleObject
cn: example backup
nsArchiveDir: /export/backups/
nsDatabaseType: ldbm database
As soon as the task is completed, the entry is removed from the directory configuration.
T he Directory Server Configuration and Command-Line Tool Reference has more information on the
available attributes for running tasks to back up databases under the cn=tasks entries.
4.3.2. Backing up the dse.ldif Configuration File
Directory Server automatically backs up the dse.ldif configuration file. When the Directory Server is
started, the directory creates a backup of the dse.ldif file automatically in a file named
dse.ldif.startOK in the /etc/dirsrv/slapd-instance_name directory.
When the dse.ldif file is modified, the file is first backed up to a file called dse.ldif.bak in the
/etc/dirsrv/slapd-instance_name directory before the directory writes the modifications to the
155
Red Hat D irectory Server 8.2 Administration Guide
dse.ldif file.
4.3.3. Restoring All Databases
T he following procedures describe restoring all of the databases in the directory using the Directory
Server Console and from the command line.
IMPORTANT
While restoring databases, the server must be running. However, the databases will be
unavailable for processing operations during the restore.
T herefore, stop all replication processes before attempting to restore a database. After
the database is restored, any consumers, hubs, or multi-master peers must be reinitialized.
4 .3.3.1. Restoring All Databases from the Console
If the databases become corrupted, restore data from a previously generated backup using the Directory
Server Console. T his process consists of stopping the server and then copying the databases and
associated index files from the backup location to the database directory.
WARNING
Restoring databases overwrites any existing database files.
IMPORTANT
While restoring databases, the server must be running. However, the databases will be
unavailable for processing operations during the restore.
T herefore, stop all replication processes before attempting to restore a database. After
the database is restored, any consumers, hubs, or multi-master peers must be reinitialized.
T o restore databases from a previously created backup:
1. In the Directory Server Console, select the T asks tab.
2. Click Restore Directory Server.
3. Select the backup from the Available Backups list, or enter the full path to a valid backup in
the Directory text box.
156
Chapter 4. Populating D irectory D atabases
T he Available Backups list shows all backups located in the default directory,
/var/lib/dirsrv/slapd-instance_name/bak/backup_directory. backup_directory is the
directory of the most recent backup, in the form serverID-YYYY_MM_DD_hhmmss.
4 .3.3.2. Restoring Databases from the Command Line
T here are three ways to restore databases from the command line:
Using the bak2db command-line script. T his script requires the server to be shut down.
Using the bak2db.pl Perl script. T his script works while the server is running.
Creating a temporary entry under cn=restore,cn=tasks,cn=config. T his method can also be
run while the server is running.
IMPORTANT
While restoring databases, the server must be running (with the exception of running the bak2db
command-line script). However, the databases will be unavailable for processing operations
during the restore.
T herefore, stop all replication processes before attempting to restore a database. After
the database is restored, any consumers, hubs, or multi-master peers must be reinitialized.
4 .3.3.2.1. Using the bak2db Command-Line Script
1. If the Directory Server is running, stop it:
service dirsrv stop instance
2. Open the Directory Server instance directory:
cd /etc/dirsrv/slapd-instance_name
3. Run the bak2db command-line script. T he bak2db script requires the full path and name of the
input file.
bak2db /var/lib/dirsrv/slapd-instance_name/bak/instance_name2010_04_30_11_48_30
For more information about using this script, see the Directory Server Configuration and
Command-Line Tool Reference.
4 .3.3.2.2. Using bak2db.pl Perl Script
157
Red Hat D irectory Server 8.2 Administration Guide
1. Open the Directory Server instance directory. For example:
cd /usr/lib[64]/dirsrv/slapd-example
2. Run the bak2db.pl Perl script.
./bak2db.pl -D "cn=Directory Manager" -w secret -a
/var/lib/dirsrv/slapd-instance_name/bak/instance_name-2010_04_30_11_48_30
For more information on using this Perl script, see the Directory Server Configuration and
Command-Line Tool Reference.
IMPORTANT
While restoring databases, the server must be running. However, the databases will be
unavailable for processing operations during the restore.
T herefore, stop all replication processes before attempting to restore a database. After
the database is restored, any consumers, hubs, or multi-master peers must be reinitialized.
Option
Description
-a
Defines the full path and name of the input file.
-D
Specifies the DN of the administrative user.
-w
Specifies the password of the administrative user.
4 .3.3.2.3. Restoring the Database through the cn=tasks Entry
T he cn=tasks,cn=config entry in the Directory Server configuration is a container entry for
temporary entries that the server uses to manage tasks. Several common directory tasks have container
entries under cn=tasks,cn=config. T emporary task entries can be created under
cn=restore,cn=tasks,cn=config to initiate a restore operation.
IMPORTANT
While restoring databases, the server must be running. However, the databases will be
unavailable for processing operations during the restore.
T herefore, stop all replication processes before attempting to restore a database. After
the database is restored, any consumers, hubs, or multi-master peers must be reinitialized.
T he restore task entry requires three attributes, the same as the backup task:
A unique name (cn).
T he directory from which to retrieve the backup file (nsArchiveDir).
T he type of database (nsDatabaseType); the only option is ldbm database.
T he entry is simply added using ldapm odify, as described in Section 3.2.4, “Adding and Modifying
Entries Using ldapmodify”. For example:
158
Chapter 4. Populating D irectory D atabases
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=example restore,cn=restore,cn=tasks,cn=config
objectclass: extensibleObject
cn: example restore
nsArchiveDir: /export/backups/
nsDatabaseType: ldbm database
As soon as the task is completed, the entry is removed from the directory configuration.
T he Directory Server Configuration and Command-Line Tool Reference has more information on the
available attributes for running database restore tasks under the cn=tasks entries.
4.3.4. Restoring a Single Database
It is possible to restore a single database through the command line, but not in the Directory Server
Console. T o restore a single database:
1. Stop the Directory Server if it is running.
service dirsrv stop instance
2. Restore the backend from the /var/lib/dirsrv/slapd-instance_name/bak archives with
the bak2db script, using the -n parameter to specify the database name. For example:
bak2db /var/lib/dirsrv/slapd-instance_name/bak/backup_file -n userRoot
3. Restart the Directory Server.
service dirsrv start instance
NOTE
If the Directory Server fails to start, remove the database transaction log files in
/var/lib/dirsrv/slapd-instance_name/db/log.###, then retry starting the
server.
4.3.5. Restoring Databases That Include Replicated Entries
If a database that supplies entries to other servers is restored, then you must reinitialize all of the
servers that receive updates from the restored database (for example, consumer servers, hub servers,
and, in multi-master replication environments, other supplier servers). T he changelog associated with
the restored database will be erased during the restore operation. A message will be logged to the
supplier servers' log files indicating that reinitialization is required. If a database containing data received
from a supplier server is restored, then one of two situations can occur:
Changelog entries have not yet expired on the supplier server.
If the supplier's changelog has not expired since the database backup was taken, then restore the
local consumer and continue with normal operations. T his situation occurs only if the backup was
taken within a period of time that is shorter than the value set for the maximum changelog age
attribute, nsslapd-changelogmaxage, in the cn=changelog5,cn=config entry. For more
information about this option, see the Directory Server Configuration and Command-Line Tool
159
Red Hat D irectory Server 8.2 Administration Guide
Reference.
Directory Server automatically detects the compatibility between the replica and its changelog. If a
mismatch is detected, the server removes the old changelog file and creates a new, empty one.
Changelog entries have expired on the supplier server since the time of the local backup.
If changelog entries have expired, reinitialize the consumer. For more information on reinitializing
consumers, refer to Section 9.10, “Initializing Consumers”.
For information on managing replication, see Chapter 9, Managing Replication.
4.3.6. Restoring the dse.ldif Configuration File
T he directory creates two backup copies of the dse.ldif file in the
/etc/dirsrv/slapd-instance_name directory. T he dse.ldif.startOK file records a copy of the
dse.ldif file at server start up. T he dse.ldif.bak file contains a backup of the most recent
changes to the dse.ldif file. Use the version with the most recent changes to restore the directory.
T o restore the dse.ldif configuration file:
1. Stop the server.
service dirsrv stop instance
2. Restore the database as outlined in Section 4.3.4, “Restoring a Single Database” to copy the
backup copy of the dse.ldif file into the directory.
3. Restart the server.
service dirsrv restart instance
160
Chapter 5. Managing Attributes and Values
Chapter 5. Managing Attributes and Values
Red Hat Directory Server provides several different mechanisms for dynamically and automatically
maintaining some types of attributes on directory entries. T hese plug-ins and configuration options
simplify managing directory data and expressing relationships between entries.
Part of the characteristic of entries are their relationships to each other. Obviously, a manager has an
employee, so those two entries are related. Groups are associated with their members. T here are less
apparent relationships, too, like between entries which share a common physical location.
Red Hat Directory Server provides several different ways that these relationshipts between entries can
be maintained smoothly and consistently. T here are several plug-ins can apply or generate attributes
automatically as part of the data within the directory, including classes of service, linking attributes, and
generating unique numeric attribute values.
5.1. Enforcing Attribute Uniqueness
Attribute uniqueness means that the Directory Server requires that all new or edited attributes always
have unique values. Attribute uniqueness is enforced through a plug-in. A new instance of the Attribute
Uniqueness Plug-in must be created for every attribute for which values must be unique.
5.1.1. Attribute Uniqueness Plug-in Syntax
Configuration information for the Attribute Uniqueness Plug-in is specified in an entry under
cn=plugins,cn=config entry. T here are two possible syntaxes for nsslapd-pluginarg attributes.
NOTE
T o enforce uniqueness of another attribute than the ones in these example, copy and paste the
default Attribute Uniqueness Plug-in entry, and being care to change only the attributes described
here.
Use the following syntax to perform the uniqueness check under a suffix or subtree:
dn: cn=descriptive_plugin_name,cn=plugins,cn=config
...
nsslapd-pluginEnabled: state
nsslapd-pluginarg0: attribute_name
nsslapd-pluginarg1: dn1
nsslapd-pluginarg2: dn2
...
Any value can be given to the cn attribute to name the plug-in. T he name should be descriptive.
T he cn attribute does not contain the name of the attribute which is checked for uniqueness.
Only one attribute can be specified on which the uniqueness check will be performed.
It is possible to specify several DNs of suffixes or subtrees in which to perform the uniqueness
check by incrementing the nsslapd-pluginarg attribute suffix by one each time.
T he variable components of the Attribute Uniqueness Plug-in syntax are described in T able 5.1,
“Attribute Uniqueness Plug-in Variables”.
Use the following syntax to specify to perform the uniqueness check below an entry containing a
specified object class:
161
Red Hat D irectory Server 8.2 Administration Guide
dn: cn=descriptive_plugin_name,cn=plugins,cn=config
...
nsslapd-pluginEnabled: state
nsslapd-pluginarg0: attribute=attribute_name
nsslapd-pluginarg1: markerObjectClass=objectclass1
nsslapd-pluginarg2: requiredObjectClass=objectclass2
...
Any value can be given to the cn attribute to name the plug-in. T he name should be descriptive.
T he cn attribute does not contain the name of the attribute which is checked for uniqueness.
Only one attribute can be specified on which the uniqueness check will be performed.
If the nsslapd-pluginarg0 attribute begins with attribute=attribute_name, then the server
expects the nsslapd-pluginarg1 attribute to include a m arkerObjectClass value.
T he variable components of the attribute uniqueness plug-in syntax are described in T able 5.1,
“Attribute Uniqueness Plug-in Variables”.
T able 5.1. Attribute Uniqueness Plug-in Variables
Variable
Definition
descriptive_plugin_name
Specifies the name of this instance of the
Attribute Uniqueness Plug-in. It is not required
that the name of the attribute for which to ensure
uniqueness be included, but it is advisable. For
example, cn=m ail
uniqueness,cn=plugins,cn=config.
state
Defines whether the plug-in is enabled or
disabled. Acceptable values are on or off.
attribute_name
T he name of the attribute for which to ensure
unique values. Only one attribute can be named.
dn
T he DN of the suffix or subtree in which to
ensure attribute uniqueness. T o specify several
suffixes or subtrees, increment the suffix of the
nsslapd-pluginarg attribute by one for each
additional suffix or subtree.
attribute= attribute_name
T he name of the attribute for which to ensure
unique values. Only one attribute can be named.
m arkerObjectClass= objectclass1
Attribute uniqueness will be checked under the
entry belonging to the DN of the updated entry
that has the object class specified in the
m arkerObjectClass keyword. Do not include a
space before or after the equals sign.
requiredObjectClass= objectclass2
Optional. When using the m arkerObjectClass
keyword to specify the scope of the uniqueness
check instead of a DN, it is also possible to
specify to perform the check only if the updated
entry contains the objectclass specified in the
requiredObjectClass keyword. Do not
include a space before or after the equals sign.
162
Chapter 5. Managing Attributes and Values
5.1.2. Creating an Instance of the Attribute Uniqueness Plug-in
T o ensure that a particular attribute in the directory always has unique values, create an instance of the
Attribute Uniqueness Plug-in for the attribute to check. For example, to ensure that every entry in the
directory that includes a mail attribute has a unique value for that attribute, create a mail uniqueness
plug-in.
T o create an instance of the Attribute Uniqueness Plug-in, modify the Directory Server configuration to
add an entry for the new plug-in under the cn=plugins,cn=config entry. T he format of the new
entry must conform to the syntax described in Section 5.1.1, “Attribute Uniqueness Plug-in Syntax”.
NOTE
Red Hat strongly encourages you to copy and paste an existing Attribute Uniqueness Plug-in
entry and only modify the attributes listed below.
For example, to create an instance the Attribute Uniqueness Plug-in for the mail attribute:
1. Stop the Directory Server. Changes to the dse.ldif file are not saved if it is edited while the
server is running.
service dirsrv stop instance_name
2. In the dse.ldif file, locate the entry for the Attribute Uniqueness Plug-in, cn=attribute
uniqueness,cn=plugins,cn=config.
3. Copy the entire entry. T he entry ends in an empty line; copy the empty line, too.
4. Paste the copied Attribute Uniqueness Plug-in entry at the end of the file.
5. Modify the Attribute Uniqueness Plug-in entry attributes for the new attribute information:
dn: cn=mail uniqueness,cn=plugins,cn=config
...
nsslapd-pluginEnabled: on
nsslapd-pluginarg0: mail
nsslapd-pluginarg1: dc=example,dc=com
...
6. Restart the Directory Server.
service dirsrv start instance_name
In this example, the uniqueness check will be performed on every entry in the dc=exam ple,dc=com
entry that includes the m ail attribute.
5.1.3. Configuring Attribute Uniqueness
T his section explains how to use Directory Server Console to view the plug-ins configured for the
directory and how to modify the configuration of the Attribute Uniqueness Plug-ins.
5.1.3.1. Configuring Attribute Uniqueness Plug-ins from the Directory Server Console
1. In the Directory Server Console, select the Configuration tab; then, in the navigation tree,
expand the Plug-ins folder, and select the Attribute Uniqueness Plug-in to modify.
163
Red Hat D irectory Server 8.2 Administration Guide
T he configuration parameters for the plug-in are displayed in the right pane.
2. T o add a suffix or subtree, click Add, and type a DN in the blank text field.
T o delete a suffix from the list, place the cursor in the text field to delete, and click Delete.
T o avoid using a DN, enter the m arkerObjectClass keyword. With this syntax, it is possible to
click Add again to specify a requiredObjectClass, as described in Section 5.1.1, “Attribute
Uniqueness Plug-in Syntax”.
NOTE
Do not add an attribute name to the list. T o check the uniqueness of other attributes, create
a new instance of the Attribute Uniqueness Plug-in for the attribute to check. For
information, see Section 5.1.2, “Creating an Instance of the Attribute Uniqueness Plug-in”.
3. Click Save.
5.1.3.2. Configuring Attribute Uniqueness Plug-ins from the Command Line
T his section provides information about configuring the plug-in from the command line.
Section 5.1.3.2.1, “Specifying a Suffix or Subtree”
Section 5.1.3.2.2, “Using the markerObjectClass and requiredObjectClass Keywords”
5.1.3.2.1. Specifying a Suffix or Subtree
T he suffix or subtrees which the plug-in checks to ensure attribute uniqueness are defined using the
nsslapd-pluginarg attribute in the entry defining the plug-in.
T o specify the subtree or subtrees, use ldapm odify to send LDIF update statements, similar to this
example:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=mail uniqueness,cn=plugins,cn=config
changetype: modify
add: nsslapd-pluginarg2 nsslapd-pluginarg3
nsslapd-pluginarg2: ou=Engineering,dc=example,dc=com
nsslapd-pluginarg3: ou=Sales,dc=example,dc=com
T his example LDIF statement modified the Attribute Uniqueness Plug-in to check the uniqueness of the
mail attribute under the subtrees dc=exam ple,dc=com ,
ou=Engineering,dc=exam ple,dc=com , and ou=Sales,dc=exam ple,dc=com .
Use the ldapm odify command to import the LDIF file into the directory. For detailed information on the
ldapm odify command, see the Directory Server Configuration and Command-Line Tool Reference.
Whenever this type of configuration change is made, restart the server.
service dirsrv restart instance_name
For information on restarting the server, see Section 1.4, “Starting and Stopping Servers”.
5.1.3.2.2. Using the markerObjectClass and requiredObjectClass Keywords
164
Chapter 5. Managing Attributes and Values
Instead of specifying a suffix or subtree in the configuration of an Attribute Uniqueness Plug-in, perform
the check under the entry belonging to the DN of the updated entry that has the object class given in the
m arkerObjectClass keyword.
T o specify to perform the uniqueness check under the entry in the DN of the updated entry that contains
the organizational unit (ou) object class, copy and paste an existing Attribute Uniqueness Plug-in entry,
and change the following attributes:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=mail uniqueness,cn=plugins,cn=config
...
nsslapd-pluginEnabled: on
nsslapd-pluginarg0: attribute=mail
nsslapd-pluginarg1: markerObjectClass=organizationalUnit
...
If the server should not check every entry in the organization unit, limit the scope by setting the check to
be performed only if the updated entry contains a specified object class.
For example, if the uniqueness of the mail attribute is checked, it is probably only necessary to perform
the check when adding or modifying entries with the person or inetorgperson object class.
Restrict the scope of the check by using the requiredObjectClass keyword, as shown in the
following example:
dn: cn=mail uniqueness,cn=plugins,cn=config
...
nsslapd-pluginEnabled: on
nsslapd-pluginarg0: attribute=mail
nsslapd-pluginarg1: markerObjectClass=organizationalUnit
nsslapd-pluginarg2: requiredObjectClass=person
...
T he m arkerObjectClass or requiredObjectClass keywords cannot be repeated by
incrementing the counter in the nsslapd-pluginarg attribute suffix. T hese keywords can only be
used once per Attribute Uniqueness Plug-in instance.
NOTE
T he nsslapd-pluginarg0 attribute always contains the name of the attribute for which to
ensure uniqueness.
5.1.4. Attribute Uniqueness Plug-in Syntax Examples
T his section contains examples of Attribute Uniqueness Plug-in syntax in the dse.ldif file.
Section 5.1.4.1, “Specifying One Attribute and One Subtree”
Section 5.1.4.2, “Specifying One Attribute and Multiple Subtrees”
5.1.4 .1. Specifying One Attribute and One Subtree
T his example configures the plug-in to ensure the uniqueness of the m ail attribute under the
165
Red Hat D irectory Server 8.2 Administration Guide
dc=exam ple,dc=com subtree.
dn: cn=mail uniqueness,cn=plugins,cn=config
...
nsslapd-pluginEnabled: on
nsslapd-pluginarg0: mail
nsslapd-pluginarg1: dc=example,dc=com
...
5.1.4 .2. Specifying One Attribute and Multiple Subtrees
It is possible use a single plug-in instance to check for the uniqueness of an attribute within multiple
subtrees, which means that the attribute value must be unique within each subtree but not unique
across all subtrees. T his example configures the Attribute Uniqueness Plug-in to ensure the
uniqueness of the mail attribute for separate subtrees, l=Chicago,dc=exam ple,dc=com and
l=Boston,dc=exam ple,dc=com .
dn: cn=mail uniqueness,cn=plugins,cn=config
...
nsslapd-pluginEnabled: on
nsslapd-pluginarg0: mail
nsslapd-pluginarg1: l=Chicago,dc=example,dc=com
nsslapd-pluginarg2: l=Boston,dc=example,dc=com
...
NOTE
T he nsslapd-pluginarg0 attribute always contains the name of the attribute for which to
ensure uniqueness. All other occurrences of the nsslapd-pluginarg, such as nsslapdpluginarg1, contain DNs.
With this configuration, the plug-in allows an instance of a value for the mail attribute to exist once
under the l=Chicago,dc=exam ple,dc=com subtree and once under the
l=Boston,dc=exam ple,dc=com subtree. For example, the following two attribute-value settings are
allowed:
mail=bjensen,l=Chicago,dc=example,dc=com
mail=bjensen,l=Boston,dc=example,dc=com
T o ensure that only one instance of a value exists under both subtrees, configure the plug-in to ensure
uniqueness for the entire dc=exam ple,dc=com subtree.
5.2. Assigning Class of Service
A class of service definition (CoS) shares attributes between entries in a way that is transparent to
applications. CoS simplifies entry management and reduces storage requirements.
Clients of the Directory Server read the attributes in a user's entry. With CoS, some attribute values may
not be stored within the entry itself. Instead, these attribute values are generated by class of service
logic as the entry is sent to the client application.
Each CoS is comprised of two types of entry in the directory:
166
Chapter 5. Managing Attributes and Values
CoS definition entry. T he CoS definition entry identifies the type of CoS used. Like the role definition
entry, it inherits from the LDAPsubentry object class. T he CoS definition entry is below the branch
at which it is effective.
Template entry. T he CoS template entry contains a list of the shared attribute values. Changes to
the template entry attribute values are automatically applied to all the entries within the scope of the
CoS. A single CoS might have more than one template entry associated with it.
T he CoS definition entry and template entry interact to provide attribute information to their target entries,
any entry within the scope of the CoS.
5.2.1. About the CoS Definition Entry
T he CoS definition entry is an instance of the cosSuperDefinition object class. T he CoS definition
entry also contains one of three object class that specifies the type of template entry it uses to generate
the entry. T he target entries which interact with the CoS share the same parent as the CoS definition
entry.
T here are three types of CoS, defined using three types of CoS definition entries:
Pointer CoS. A pointer CoS identifies the template entry using the template DN only.
Indirect CoS. An indirect CoS identifies the template entry using the value of one of the target entry's
attributes. For example, an indirect CoS might specify the manager attribute of a target entry. T he
value of the manager attribute is then used to identify the template entry.
T he target entry's attribute must be single-valued and contain a DN.
Classic CoS. A classic CoS identifies the template entry using a combination of the template entry's
base DN and the value of one of the target entry's attributes.
For more information about the object classes and attributes associated with each type of CoS, refer to
Section 5.2.11, “Managing CoS from the Command Line”.
If the CoS logic detects that an entry contains an attribute for which the CoS is generating values, the
CoS, by default, supplies the client application with the attribute value in the entry itself. However, the
CoS definition entry can control this behavior.
5.2.2. About the CoS Template Entry
T he CoS template entry contains the value or values of the attributes generated by the CoS logic. T he
CoS template entry contains a general object class of cosT em plate. T he CoS template entries for a
given CoS are stored in the directory tree along with the CoS definition.
T he relative distinguished name (RDN) of the template entry is determined by one of the following:
T he DN of the template entry alone. T his type of template is associated with a pointer CoS definition.
T he value of one of the target entry's attributes. T he attribute used to provide the relative DN to the
template entry is specified in the CoS definition entry using the cosIndirectSpecifier attribute.
T his type of template is associated with an indirect CoS definition.
By a combination of the DN of the subtree where the CoS performs a one level search for templates
and the value of one of the target entry's attributes. T his type of template is associated with a classic
CoS definition.
5.2.3. How a Pointer CoS Works
An administrator creates a pointer CoS that shares a common postal code with all of the entries stored
under dc=exam ple,dc=com . T he three entries for this CoS appear as illustrated in Figure 5.1,
“Sample Pointer CoS”.
167
Red Hat D irectory Server 8.2 Administration Guide
Figure 5.1. Sample Pointer CoS
In this example, the template entry is identified by its DN, cn=exam pleUS,cn=data, in the CoS
definition entry. Each time the postalCode attribute is queried on the entry
cn=wholiday,ou=people,dc=exam ple,dc=com , the Directory Server returns the value available
in the template entry cn=exam pleUS,cn=data.
5.2.4. How an Indirect CoS Works
An administrator creates an indirect CoS that uses the manager attribute of the target entry to identify
the template entry. T he three CoS entries appear as illustrated in Figure 5.2, “Sample Indirect CoS”.
Figure 5.2. Sample Indirect CoS
168
Chapter 5. Managing Attributes and Values
In this example, the target entry for William Holiday contains the indirect specifier, the manager attribute.
William's manager is Carla Fuentes, so the manager attribute contains a pointer to the DN of the
template entry, cn=Carla Fuentes,ou=people,dc=exam ple,dc=com . T he template entry in turn
provides the departmentNumber attribute value of 31884 2.
5.2.5. How a Classic CoS Works
An administrator creates a classic CoS that uses a combination of the template DN and a CoS specifier
to identify the template entry containing the postal code. T he three CoS entries appear as illustrated in
Figure 5.3, “Sample Classic CoS”:
Figure 5.3. Sample Classic CoS
In this example, the CoS definition entry's cosSpecifier attribute specifies the employeeType attribute.
T his attribute, in combination with the template DN, identify the template entry as
cn=sales,cn=exam pleUS,cn=data. T he template entry then provides the value of the postalCode
attribute to the target entry.
5.2.6. Handling Physical Attribute Values
T he cosAttribute attribute contains the name of another attribute which is governed by the class of
service. T his attribute allows an override qualifier (after the attribute value) which sets how the CoS
handles existing attribute values on entries when it generates attribute values.
cosAttribute: attribute_name override
T here are four override qualifiers.
Override Qualifier
Description
default
Only returns a generated value if there is no
corresponding attribute value stored with the
169
Red Hat D irectory Server 8.2 Administration Guide
entry.
override
Always returns the value generated by the CoS,
even when there is a value stored with the entry.
operational
Returns a generated attribute only if it is explicitly
requested in the search. Operational attributes do
not need to pass a schema check in order to be
returned. When operational is used, it also
overrides any existing attribute values.
NOTE
An attribute can only be made operational
if it is defined as operational in the
schema. For example, if the CoS generates
a value for the description attribute, it is
not possible to use the operational
qualifier because this attribute is not
marked operational in the schema.
operational-default
Only returns a generated value if there is no
corresponding attribute value stored with the
entry and if it is explicitly requested in the search.
If no qualifier is set, default is assumed.
For example, this pointer CoS definition entry indicates that it is associated with a template entry,
cn=exam pleUS,ou=data,dc=exam ple,dc=com , that generates the value of the postalCode
attribute. T he override qualifier indicates that this value will take precedence over the value stored by
the entries for the postalCode attribute:
dn: cn=pointerCoS,dc=example,dc=com
objectclass: top
objectclass: cosSuperDefinition
objectclass: cosPointerDefinition
cosTemplateDn: cn=exampleUS,ou=data,dc=example,dc=com
cosAttribute: postalCode override
NOTE
If an entry contains an attribute value generated by a CoS, the value of the attribute cannot be
manually updated if it is defined with the operational or override qualifiers.
For more information about the CoS attributes, see the Directory Server Configuration and CommandLine Tool Reference.
5.2.7. Handling Multi-valued Attributes with CoS
Any attribute can be generated using a class of service — including multi-valued attributes. T hat
introduces the potential for confusion. Which CoS supplies a value? Any of them or all of them? How is
the value selected from competing CoS templates? Does the generated attribute use a single value or
multiple values?
170
Chapter 5. Managing Attributes and Values
T here way to resolve this is to set a priority to select one CoS value out of competing CoS definitions.
T his generates one single value for the target entry.
NOTE
Indirect CoS don't support the cosPriority attribute.
If there are multiple CoS definitions available for an attribute, then the value is chosen arbitrarily. T his is
an unpredictable and unwieldy option. T he way to control which CoS template to use is to set a ranking
on the template — a priority — and the highest prioritized CoS always "wins" and provides the value.
It is fairly common for there to be multiple templates completing to provide a value. For example, there
can be a multi-valued cosSpecifier attribute in the CoS definition entry. T he template priority is set
using the cosPriority attribute. T his attribute represents the global priority of a particular template. A
priority of zero is the highest priority.
For example, a CoS template entry for generating a department number appears as follows:
dn: cn=data,dc=example,dc=com
objectclass: top
objectclass: extensibleObject
objectclass: cosTemplate
departmentNumber: 71776
cosPriority: 0
T his template entry contains the value for the departmentNumber attribute. It has a priority of zero,
meaning this template takes precedence over any other conflicting templates that define a different
departmentNumber value.
T emplates that contain no cosPriority attribute are considered the lowest priority. Where two or more
templates are considered to supply an attribute value and they have the same (or no) priority, a value is
chosen arbitrarily.
NOTE
T he behavior for negative cosPriority values is not defined in Directory Server; do not enter
negative values.
5.2.8. Searches for CoS-Specified Attributes
CoS definitions provide values for attributes in entries. For example, a CoS can set the postalCode
attribute for every entry in a subtree. Searches against those CoS-defined attributes, however, do not
behave like searches against regular entries.
If the CoS-defined attribute is indexed with any kind of index (including presence), then any attribute with
a value set by the CoS is not returned with a search. For example:
T he postalCode attribute for T ed Morris is defined by a CoS.
T he postalCode attribute for Barbara Jensen is set in her entry.
T he postalCode attribute is indexed.
171
Red Hat D irectory Server 8.2 Administration Guide
If an ldapsearch command uses the filter (postalCode=* ), then Barbara Jensen's entry is
returned, while T ed Morris's is not.
If the CoS-defined attribute is not indexed, then every matching entry is returned in a search, regardless
of whether the attribute value is set locally or with CoS. For example:
T he postalCode attribute for T ed Morris is defined by a CoS.
T he postalCode attribute for Barbara Jensen is set in her entry.
T he postalCode attribute is not indexed.
If an ldapsearch command uses the filter (postalCode=* ), then both Barbara Jensen's and T ed
Morris's entries are returned.
CoS allows for an override, an identifier given to the cosAttribute attribute in the CoS entry, which
means that local values for an attribute can override the CoS value. If an override is set on the CoS, then
an ldapsearch operation will return a value for an entry even if the attribute is indexed, as long as
there is a local value for the entry. Other entries which possess the CoS but do not have a local value
will still not be returned in the ldapsearch operation.
Because of the potential issues with running LDAP search requests on CoS-defined attributes, take care
when deciding which attributes to generate using a CoS.
5.2.9. Access Control and CoS
T he server controls access to attributes generated by a CoS in exactly the same way as regular stored
attributes. However, access control rules depending upon the value of attributes generated by CoS will
not work. T his is the same restriction that applies to using CoS-generated attributes in search filters.
5.2.10. Managing CoS Using the Console
T his section describes creating and editing CoS through the Directory Server Console:
Section 5.2.10.1, “Creating a New CoS”
Section 5.2.10.2, “Creating the CoS T emplate Entry”
5.2.10.1. Creating a New CoS
1. In the Directory Server Console, select the Directory tab.
2. Browse the tree in the left navigation pane, and select the parent entry for the new class of
service.
3. Go to the Object menu, and select New > Class of Service.
172
Chapter 5. Managing Attributes and Values
Alternatively, right-click the entry and select New > Class of Service.
4. Select General in the left pane. In the right pane, enter the name of the new class of service in
the Class Nam e field. Enter a description of the class in the Description field.
5. Click Attributes in the left pane. T he right pane displays a list of attributes generated on the
target entries.
Click Add to browse the list of possible attributes and add them to the list.
173
Red Hat D irectory Server 8.2 Administration Guide
6. After an attribute is added to the list, a drop-down list appears in the Class of Service
Behavior column.
Select Does not override target entry attribute to tell the directory to only return
a generated value if there is no corresponding attribute value stored with the entry.
Select Overrides target entry attribute to make the value of the attribute generated
by the CoS override the local value.
Select Overrides target entry attribute and is operational to make the
attribute override the local value and to make the attribute operational, so that it is not visible to
174
Chapter 5. Managing Attributes and Values
client applications unless explicitly requested.
Select Does not override target entry attribute and is operational to tell
the directory to return a generated value only if there is no corresponding attribute value
stored with the entry and to make the attribute operational (so that it is not visible to client
applications unless explicitly requested).
NOTE
An attribute can only be made operational if it is also defined as operational in the schema.
For example, if a CoS generates a value for the description attribute, you cannot select
Overrides target entry attribute and is operational because this
attribute is not marked operational in the schema.
7. Click T em plate in the left pane. In the right pane, select how the template entry is identified.
By its DN. T o have the template entry identified by only its DN (a pointer CoS), enter the DN of
the template in the T em plate DN field. Click Browse to locate the DN on the local server.
T his will be an exact DN, such as cn=CoS tem plate,ou=People,dc=exam ple,dc=com .
Using the value of one of the target entry's attribute. T o have the template entry identified by
the value of one of the target entry's attributes (an indirect CoS), enter the attribute name in
the Attribute Nam e field. Click Change to select a different attribute from the list of
available attributes.
Using both its DN and the value of one of the target entry's attributes. T o have the template
entry identified by both its DN and the value of one of the target entry's attributes (a classic
CoS), enter both a template DN and an attribute name. T he template DN in a classic CoS is
more general than for a pointer CoS; it references the suffix or subsuffix where the template
entries will be. T here can be more than one template for a classic CoS.
8. Click OK.
5.2.10.2. Creating the CoS T emplate Entry
175
Red Hat D irectory Server 8.2 Administration Guide
For a pointer CoS or a classic CoS, there must be a template entry, according to the template DN set
when the class of service was created. Although the template entries can be placed anywhere in the
directory as long as the cosTemplateDn attribute reflects that DN, it is best to place the template entries
under the CoS itself.
For a pointer CoS, make sure that this entry reflects the exact DN given when the CoS was created.
For a classic CoS, the template DN should be recursive, pointing back to the CoS entry itself as the
base suffix for the template.
1. In the Directory Server Console, select the Directory tab.
2. Browse the tree in the left navigation pane, and select the parent entry that contains the class of
service.
T he CoS appears in the right pane with other entries.
3. Right-click the CoS, and select New > Other.
Alternatively, select the CoS in the right pane, click Object in the menu at the top, and select New
> Other.
176
Chapter 5. Managing Attributes and Values
4. Select cosT em plate from the list of object classes.
NOTE
T he LDAPsubentry object class can be added to a new template entry. Making the CoS
template entry an instance of the LDAPsubentry object class allows ordinary searches to
be performed unhindered by the configuration entries. However, if the template entry
already exists and is used for something else (for example, if it is a user entry), the
LDAPsubentry object class does not need to be added to the template entry.
5. Select the object classes attribute, and click Add Value.
177
Red Hat D irectory Server 8.2 Administration Guide
6. Add the extensibleObject object class. T his makes it possible to add any attribute available
in the directory.
7. Click the Add Attribute button.
178
Chapter 5. Managing Attributes and Values
8. Add the cn attribute, and give it a value that corresponds to the attribute value in the target entry.
For example, if the manager attribute is used to set the value for a classic CoS, give the cn a value
of a manager's DN, such as uid=bparker,ou=people,dc=exam ple,dc=com . Alternatively,
set it to a role, such as cn=QA Role,dc=exam ple,dc=com or a regular attribute value. For
example, if the employeeType attribute is selected, it can be full tim e or tem porary.
9. Click the Change button in the lower right corner to change the naming attribute.
179
Red Hat D irectory Server 8.2 Administration Guide
10. Use the cn of the entry as the naming attribute instead of cosPriority.
11. Click the Add Attribute button, and add the attributes listed in the CoS. T he values used here
will be used throughout the directory in the targeted entries.
12. Set the cosPriority. T here may be more than one CoS that applies to a given attribute in an
entry; the cosPriority attribute ranks the importance of that particular CoS. T he higher
cosPriority will take precedence in a conflict. T he highest priority is 0.
T emplates that contain no cosPriority attribute are considered the lowest priority. In the case
where two or more templates could supply an attribute value and they have the same (or no)
priority, a value is chosen arbitrarily.
NOTE
T he behavior for negative cosPriority values is not defined in Directory Server; do not
enter negative values.
180
Chapter 5. Managing Attributes and Values
NOTE
T he cosPriority attribute is not supported by indirect CoS.
T he CoS is visible in the left navigation pane once there are entries beneath it. For classic CoS, there
can be multiple entries, according to the different potential values of the attribute specifier.
T o edit the description or attributes generated on the target entry of an existing CoS, simply double-click
the CoS entry listed in the Directory tab, and make the appropriate changes in the editor window.
5.2.11. Managing CoS from the Command Line
Because all configuration information and template data is stored as entries in the directory, standard
LDAP tools can be used for CoS configuration and management.
Section 5.2.11.1, “Creating the CoS Definition Entry from the Command Line”
Section 5.2.11.2, “Creating the CoS T emplate Entry from the Command Line”
Section 5.2.11.3, “Example of a Pointer CoS”
Section 5.2.11.4, “Example of an Indirect CoS”
Section 5.2.11.5, “Example of a Classic CoS”
Section 5.2.11.6, “Searching for CoS Entries”
5.2.11.1. Creating the CoS Definition Entry from the Command Line
Each type of CoS requires a particular object class to be specified in the definition entry. All CoS
definition object classes inherit from the LDAPsubentry object class and the cosSuperDefinition
object class.
A pointer CoS uses the cosPointerDefinition object class. T his object class identifies the
template entry using an entry DN value specified in the cosTemplateDn attribute, as shown in
Example 5.1, “An Example Pointer CoS Entry”.
Example 5.1. An Example Pointer CoS Entry
dn: cn=pointerCoS,dc=example,dc=com
objectclass: top
objectclass: cosSuperDefinition
objectclass: cosPointerDefinition
cosTemplateDn:DN_string
cosAttribute:list_of_attributes qualifier
cn: pointerCoS
An indirect CoS uses the cosIndirectDefinition object class. T his type of CoS identifies the
template entry based on the value of one of the target entry's attributes, as specified in the
cosIndirectSpecifier attribute. T his is illustrated in Example 5.2, “An Example Indirect CoS Entry”.
181
Red Hat D irectory Server 8.2 Administration Guide
Example 5.2. An Example Indirect CoS Entry
dn: cn=indirectCoS,dc=example,dc=com
objectclass: top
objectclass: cosSuperDefinition
objectclass: cosIndirectDefinition
cosIndirectSpecifier:attribute_name
cosAttribute:list_of_attributes qualifier
cn: indirectCoS
A classic CoS uses the cosClassicDefinition object class. T his identifies the template entry using
both the template entry's DN (set in the cosTemplateDn attribute) and the value of one of the target
entry's attributes (set in the cosSpecifier attribute). T his is illustrated in Example 5.3, “An Example
Classic CoS Entry”.
Example 5.3. An Example Classic CoS Entry
dn: cn=classicCoS,dc=example,dc=com
objectclass: top
objectclass: cosSuperDefinition
objectclass: cosClassicDefinition
cosTemplateDn:DN_string
cosSpecifier:attribute_name
cosAttribute:list_of_attributes qualifier
cn: classicCoS
For a class of service, the object class defines the type of CoS, and the supporting attributes identify
which directory entries are affected by defining the CoS template. Every CoS has one additional attribute
which can be defined for it: cosAttribute. T he purpose of a CoS is to supply attribute values across
multiple entries; the cosAttribute attribute defines which attribute the CoS generates values for.
5.2.11.2. Creating the CoS T emplate Entry from the Command Line
Each template entry is an instance of the cosT em plate object class.
NOTE
Consider adding the LDAPsubentry object class to a new template entry. Making the CoS
template entry an instance of the LDAPsubentry object classes allows ordinary searches to be
performed unhindered by the configuration entries. However, if the template entry already exists
and is used for something else, such as a user entry, the LDAPsubentry object class does not
need to be added to the template entry.
T he CoS template entry also contains the attribute generated by the CoS (as specified in the
cosAttribute attribute of the CoS definition entry) and the value for that attribute.
For example, a CoS template entry that provides a value for the postalCode attribute follows:
182
Chapter 5. Managing Attributes and Values
dn:cn=exampleUS,ou=data,dc=example,dc=com
objectclass: top
objectclass: extensibleObject
objectclass: cosTemplate
postalCode: 44438
T he following sections provide examples of template entries along with examples of each type of CoS
definition entry.
Section 5.2.11.3, “Example of a Pointer CoS”
Section 5.2.11.4, “Example of an Indirect CoS”
Section 5.2.11.5, “Example of a Classic CoS”
5.2.11.3. Example of a Pointer CoS
Example Corporation's administrator is creating a pointer CoS that shares a common postal code with all
entries in the dc=exam ple,dc=com tree.
1. Add a new pointer CoS definition entry to the dc=exam ple,dc=com suffix using ldapm odify:
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389
-h server.example.com
T he ldapm odify utility binds to the server and prepares it to add information to the
configuration file.
2. Next, add the pointer CoS definition to the dc=exam ple,dc=com root suffix.
dn: cn=pointerCoS,dc=example,dc=com
objectclass: top
objectclass: cosSuperDefinition
objectclass: cosPointerDefinition
cosTemplateDn: cn=exampleUS,ou=data,dc=example,dc=com
cosAttribute: postalCode
3. Create the template entry.
dn: cn=exampleUS,ou=data,dc=example,dc=com
objectclass: top
objectclass: extensibleObject
objectclass: cosTemplate
postalCode: 44438
T he CoS template entry (cn=exam pleUS,ou=data,dc=exam ple,dc=com ) supplies the value
stored in its postalCode attribute to any entries located under the dc=exam ple,dc=com suffix.
T hese entries are the target entries.
5.2.11.4 . Example of an Indirect CoS
T his indirect CoS uses the manager attribute of the target entry to identify the CoS template entry, which
varies depending on the different values of the attribute.
1. Add a new indirect CoS definition entry to the dc=exam ple,dc=com suffix, using ldapm odify
as follows:
183
Red Hat D irectory Server 8.2 Administration Guide
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389
-h server.example.com
T he ldapm odify utility binds to the server and prepares it to add information to the
configuration file.
2. Add the indirect CoS definition to the dc=exam ple,dc=com root suffix as follows:
dn: cn=indirectCoS,dc=example,dc=com
objectclass: top
objectclass: cosSuperDefinition
objectclass: cosIndirectDefinition
cosIndirectSpecifier: manager
cosAttribute: departmentNumber
If the directory or modify the manager entries already contain the departmentNumber attribute, then no
other attribute needs to be added to the manager entries. T he definition entry looks in the target suffix
(the entries under dc=exam ple,dc=com ) for entries containing the manager attribute because this
attribute is specified in the cosIndirectSpecifier attribute of the definition entry). It then checks the
departmentNumber value in the manager entry that is listed. T he value of the departmentNumber
attribute will automatically be relayed to all of the manager's subordinates that have the manager
attribute. T he value of departmentNumber will vary depending on the department number listed in the
different manager's entries.
5.2.11.5. Example of a Classic CoS
T he Example Corporation administrator is creating a classic CoS that automatically generates postal
codes using a combination of the template DN and the attribute specified in the cosSpecifier attribute.
1. Add a new classic CoS definition entry to the dc=exam ple,dc=com suffix, using ldapm odify.
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389
-h server.example.com
T he ldapm odify utility binds to the server and prepares it to add information to the
configuration file.
2. Add the indirect CoS definition to the dc=exam ple,dc=com root suffix.
dn: cn=classicCoS,dc=example,dc=com
objectclass: top
objectclass: cosSuperDefinition
objectclass: cosClassicDefinition
cosTemplateDn: cn=classicCoS,dc=example,dc=com
cosSpecifier: businessCategory
cosAttribute: postalCode override
3. Create the template entries for the sales and marketing departments. Add the CoS attributes to
the template entry. T he cn of the template sets the value of the businessCategory attribute in
the target entry, and then the attributes are added or overwritten according to the value in the
template:
184
Chapter 5. Managing Attributes and Values
dn: cn=sales,cn=classicCoS,dc=example,dc=com
objectclass: top
objectclass: extensibleObject
objectclass: cosTemplate
postalCode: 44438
dn: cn=marketing,cn=classicCoS,dc=example,dc=com
objectclass: top
objectclass: extensibleObject
objectclass: cosTemplate
postalCode: 99111
T he classic CoS definition entry applies to all entries under the dc=exam ple,dc=com suffix.
Depending upon the combination of the businessCategory attribute found in the entry and the
cosTemplateDn, it can arrive at one of two templates. One, the sales template, provides a postal code
specific to employees in the sales department. T he marketing template provides a postal code specific
to employees in the marketing department.
5.2.11.6. Searching for CoS Entries
CoS definition entries are operational entries and are not returned by default with regular searches. T his
means that if a CoS is defined under ou=People,dc=exam ple,dc=com , for example, the following
ldapsearch command will not return them:
ldapsearch -s sub -b ou=People,dc=example,dc=com “(objectclass=*)”
T o return the CoS definition entries, add the ldapSubEntry object class to the CoS definition entries.
For example:
dn: cn=pointerCoS,ou=People,dc=example,dc=com
objectclass: top
objectclass: cosSuperDefinition
objectclass: cosPointerDefinition
objectclass: ldapSubEntry
cosTemplateDn: cn=exampleUS,ou=data,dc=example,dc=com
cosAttribute: postalCode override
T hen use a special search filter, (objectclass=ldapSubEntry), with the search. T his filter can be
added to any other search filter using OR (|):
ldapsearch -s sub -b ou=People,dc=example,dc=com
“(|(objectclass=*)(objectclass=ldapSubEntry))”
T his search returns all regular entries in addition to CoS definition entries in the
ou=People,dc=exam ple,dc=com subtree.
NOTE
T he Console automatically shows CoS entries.
5.2.12. Creating Role-Based Attributes
Classic CoS schemes generate attribute values for an entry based on the role possessed by the entry.
For example, role-based attributes can be used to set the server look-through limit on an entry-by-entry
185
Red Hat D irectory Server 8.2 Administration Guide
For example, role-based attributes can be used to set the server look-through limit on an entry-by-entry
basis.
T o create a role-based attribute, use the nsRole attribute as the cosSpecifier in the CoS definition
entry of a classic CoS. Because the nsRole attribute can be multi-valued, CoS schemes can be defined
that have more than one possible template entry. T o resolve the ambiguity of which template entry to
use, include the cosPriority attribute in the CoS template entry.
For example, this CoS allows members of the manager role to exceed the standard mailbox quota. T he
manager role entry is:
dn: cn=ManagerRole,ou=people,dc=example,dc=com
objectclass: top
objectclass: nsRoleDefinition
objectclass: nsComplexRoleDefinition
objectclass: nsFilteredRoleDefinition
cn: ManagerRole
nsRoleFilter: ou=managers
Description: filtered role for managers
IMPORTANT
T he nsRoleFilter attribute cannot accept virtual attribute values.
T he classic CoS definition entry looks like:
dn: cn=managerCOS,dc=example,dc=com
objectclass: top
objectclass: cosSuperDefinition
objectclass: cosClassicDefinition
cosTemplateDn: cn=managerCOS,dc=example,dc=com
cosSpecifier: nsRole
cosAttribute: mailboxquota override
T he cosTemplateDn attribute provides a value that, in combination with the attribute specified in the
cosSpecifier attribute (in the example, the nsRole attribute of the target entry), identifies the CoS
template entry. T he CoS template entry provides the value for the mailboxquota attribute. An additional
qualifier of override tells the CoS to override any existing mailboxquota attributes values in the
target entry.
T he corresponding CoS template entry looks as follows:
dn:cn=cn=ManagerRole\,ou=people\,dc=example\,dc=com,cn=managerCOS,dc=example,dc=c
om
objectclass: top
objectclass: extensibleObject
objectclass: cosTemplate
mailboxquota: 1000000
T he template provides the value for the mailboxquota attribute, 1000000.
186
Chapter 5. Managing Attributes and Values
NOTE
T he role entry and the CoS definition and template entries should be located at the same level in
the directory tree.
5.3. Linking Attributes to Manage Attribute Values
A class of service dynamically supplies attribute values for entries which all have attributes with the
same value, like building addresses, postal codes, or main office numbers. T hese are shared attribute
values, which are updated in a single template entry.
Frequently, though, there are relationships between entries where there needs to be a way to express
linkage between them, but the values (and possibly even the attributes) that express that relationship
are different. Red Hat Directory Server provides a way to link specified attributes together, so that when
one attribute in one entry is altered, a corresponding attribute on a related entry is automatically
updated. (T he link and managed attributes both have DN values. T he value of the link attribute contains
the DN of the entry for the plug-in to update; the managed attribute in the second entry has a DN value
which points back to the original link entry.)
5.3.1. About Linking Attributes
T he Linked Attributes Plug-in, allows multiple instances of the plug-in. Each instance configures one
attribute which is manually maintained by the administrator (linkType) and one attribute which is
automatically maintained by the plug-in (managedType).
Figure 5.4 . Basic Linked Attribute Configuration
187
Red Hat D irectory Server 8.2 Administration Guide
NOTE
T o preserve data consistency, only the plug-in process should maintain the managed attribute.
Consider creating an ACI that will restrict all write access to any managed attribute. See
Section 12.5.2, “Creating a New ACI” T o preserve data consistency, only the plug-in process
should maintain the managed attribute. Consider creating an ACI that will restrict all write access
to any managed attribute. See Section 12.5.2, “Creating a New ACI” for information on setting
ACIs.
A Linked Attribute Plug-in instance can be restricted to a single subtree within the directory. T his can
allow more flexible customization of attribute combinations and affected entries. If no scope is set, then
the plug-in operates in the entire directory.
Figure 5.5. Restricting the Linked Attribute Plug-in to a Specific Subtree
When configuring the Linked Attribute Plug-in instance, certain configurations are required:
Both the managed attribute and linked attribute must require the Distinguished Name syntax in their
attribute definitions. T he linked attributes are essentially managed cross-references, and the way
that the plug-in handles these cross-references is by pulling the DN of the entry from the attribute
value.
For information on planning custom schema elements, see Chapter 6, Managing the Directory
Schema.
Each Linked Attribute Plug-in instance must be local and any managed attributes must be blocked
from replication using fractional replication.
Any changes that are made on one supplier will automatically trigger the plug-in to manage the
values on the corresponding directory entries, so the data stay consistent across servers. However,
the managed attributes must be maintained by the plug-in instance for the data to be consistent
between the linked entries. T his means that managed attribute values should be maintained solely
by the plug-in processes, not the replication process, even in a multi-master replication environment.
For information on using fractional replication, see Section 9.1.7, “Replicating a Subset of Attributes
with Fractional Replication”.
5.3.2. Looking at the Linking Attributes Plug-in Syntax
T he default Linked Attributes Plug-in entry is a container entry for each plug-in instance, similar to the
password syntax plug-ins or the DNA Plug-in in the next section. Each entry beneath this container entry
defines a different link-managed attribute pair.
188
Chapter 5. Managing Attributes and Values
T o create a new linking attribute pair, then, create a new plug-in instance beneath the container entry. A
basic linking attribute plug-in instance required defining two things:
T he attribute that is managed manually by administrators, in the linkType attribute
T he attribute that is created dynamically by the plug-in, in the managedType attribute
Optionally, a scope that restricts the plug-in to a specific part of the directory tree, in the linkScope
attribute
Example 5.4 . Example Linked Attributes Plug-in Instance Entry
dn: cn=Manager Link,cn=Linked Attributes,cn=plugins,cn=config
objectClass: top
objectClass: extensibleObject
cn: Manager Link1
cn: Manager Link
linkType: directReport
managedType: manager
linkScope: ou=people,dc=example,dc=com
All of the attributes available for an instance of the Linked Attributes Plug-in instance are listed in
T able 5.2, “Linked Attributes Plug-in Instance Attributes”.
T able 5.2. Linked Attributes Plug-in Instance Attributes
Plug-in Attribute
Description
cn
Gives a unique name for the plug-in instance.
linkScope
Contains the DN of a suffix to which to restrict the
function of the plug-in instance.
linkT ype
Gives the attribute which is maintained by an
administrator. T his attribute is manually
maintained and is used as the reference for the
plug-in. T his attribute must have a DN value
format. When the attribute is added, modified, or
deleted, then its value contains the DN of the
target entry for the plug-in to update.
managedT ype
Gives the attribute which is maintained by the
plug-in. T his attribute is created and updated on
target entries. T his attribute must have a DN
value format. When the attribute is added to the
entry, its value will point back as a crossreference to the managed entry.
5.3.3. Configuring Attribute Links
NOTE
T he Linked Attribute Plug-in instance can be created in the Directory Server Console, but only
through the Advanced Property Editor for the directory entry, by manually adding all of the
required attributes, the same as creating the entry manually through the command line.
189
Red Hat D irectory Server 8.2 Administration Guide
1. If it is not already enabled, enable the Linked Attributes Plug-in, as described in Section 1.10.1,
“Enabling Plug-ins”. For example:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389
dn: cn=Linked Attributes,cn=plugins,cn=config
changetype: modify
replace: nsslapd-pluginEnabled
nsslapd-pluginEnabled: on
2. Create the plug-in instance. Both the managedType and linkType attributes are required. T he
plug-in syntax is covered in Section 5.3.2, “Looking at the Linking Attributes Plug-in Syntax”. For
example:
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389
-h server.example.com
dn: cn=Manager Link,cn=Linked Attributes,cn=plugins,cn=config
objectClass: top
objectClass: extensibleObject
cn: Manager Link
linkType: directReport
managedType: manager
3. Restart the server to apply the new plug-in instance.
service dirsrv restart
5.3.4. Cleaning up Attribute Links
T he managed-linked attributes can get out of sync. For instance, a linked attribute could be imported or
replicated over to a server, but the corresponding managed attribute was not because the link attribute
wasn't properly configured. T he managed-linked attribute pairs can be fixed by running a script (fixuplinkedattrs.pl) or by launching a fix-up task.
T he fixup task removes any managed attributes (attributes managed by the plug-in) that do not have a
corresponding link attribute (attributes managed by the administrator) on the referenced entry.
Conversely, the task adds any missing managed attributes if the link attribute exists in an entry.
5.3.4 .1. Regenerating Linked Attributes Using fixup-linkedattrs.pl
T he fixup-linkedattrs.pl script launches a special task to regenerate all of the managed-link
attribute pairs on directory entries. One or the other may be lost in certain situations. If the link attribute
exists in an entry, the task traces the cross-referenced DN in the available attribute and creates the
corresponding configured managed attribute on the referenced entry. If a managed attribute exists with
no corresponding link attribute, then the managed attribute value is removed.
T o repair all configured link attribute pairs for the entire scope of the plug-in, then simply run the
command as the Directory Manager:
/usr/lib64/dirsrv/instance_name/fixup-linkedattrs.pl -D "cn=Directory Manager" -w
password
It is also possible to limit the fixup task to a single link-managed attribute pair, using the -l option to
specify the target plug-in instance DN:
190
Chapter 5. Managing Attributes and Values
/usr/lib64/dirsrv/instance_name/fixup-linkedattrs.pl -D "cn=Directory Manager" -w
password -l "cn=Manager Link,cn=Linked Attributes,cn=plugins,cn=config"
T he fixup-linkedattrs.pl tool is described in more detail in the Configuration and Command-Line
Tool Reference.
5.3.4 .2. Regenerating Linked Attributes Using ldapmodify
Repairing linked attributes is one of the tasks which can be managed through a special task
configuration entry. T ask entries occur under the cn=tasks configuration entry in the dse.ldif file, so
it is also possible to initiate a task by adding the entry using ldapm odify. When the task is complete,
the entry is removed from the directory.
T his task is the same one created automatically by the fixup-linkedattrs.pl script when it is run.
T o initiate a linked attributes fixup task, add an entry under the cn=fixup linked
attributes,cn=tasks,cn=config entry. T he only required attribute is the cn for the specific task,
though it also allows the ttl attribute to set a timeout period.
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=example,cn=fixup linked attributes,cn=tasks,cn=config
cn:example
ttl: 5
Once the task is completed, the entry is deleted from the dse.ldif configuration, so it is possible to
reuse the same task entry continually.
T he cn=fixup linked attributes task configuration is described in more detail in the
Configuration and Command-Line Tool Reference.
5.4. Assigning and Managing Unique Numeric Attribute Values
Some entry attributes require having a unique number, such as uidNumber and gidNumber. T he
Directory Server can automatically generate and supply unique numbers for specified attributes using
the Distributed Numeric Assignment (DNA) Plug-in.
NOTE
Attribute uniqueness is not necessarily preserved with the DNA Plug-in. T he plug-in only assigns
non-overlapping ranges, but it does allow manually-assigned numbers for its managed attributes,
and it does not verify or require that the manually-assigned numbers are unique.
T he issue with assigning unique numbers is not with generating the numbers but in effectively managing
numbers so that there are not any conflicts with other assigned numbers when entries are replicated
and that every server has a sufficient range of numbers to assign.
T he DNA Plug-in for a server assigns a range of available numbers that that instance can issue. T he
range definition is very simple and is set by two attributes: the server's next available number (the low
end of the range) and its maximum value (the top end of the range). T he initial bottom range is set when
the plug-in instance is configured. After that, the bottom value is updated by the plug-in. By breaking the
available numbers into ranges, the servers can all continually assign numbers without overlapping with
191
Red Hat D irectory Server 8.2 Administration Guide
each other.
T he server performs a sorted search, internally, to see if the next specified range is already taken,
requiring the managed attribute to have an equality index with the proper ordering matching rule (as
described in Section 7.2, “Creating Standard Indexes”).
For multi-master replication, each supplier can be configured with a threshold, so that when it begins
running out of numbers in its range, it can request additional ranges from other suppliers. Each supplier
keeps a track of its current range in a separate configuration entry. T he configuration entry is replicated
to all of the other suppliers, so each supplier can check that configuration to find a server to contact for a
new range. For example:
dn: dnaHostname=ldap1.example.com+dnaPortNum=389,cn=Account
UIDs,ou=Ranges,dc=example,dc=com
objectClass: extensibleObject
objectClass: top
dnahostname: ldap1.example.com
dnaPortNum: 389
dnaSecurePortNum: 636
dnaRemainingValues: 1000
T he DNA Plug-in is applied, always, to a specific area of the directory tree (the scope) and to specific
entry types within that subtree (the filter).
T here are several different ways that the Directory Server can handle generating attribute values.
In the simplest case, a user entry is added to the directory with an object class which requires the
unique-number attribute, but without the attribute. Adding (or requiring) the managed attribute without a
value triggers the DNA Plug-in to assign a value.
A similar and more manageable option is to use a magic number. T his magic number is a template value
for the managed attribute, something outside the server's range, a number or even a word, that the plugin recognizes it needs to replace with a new assigned value. When an entry is added with that number,
and the entry is within the scope and filter of the configured DNA Plug-in, then using the magic number
automatically triggers the plug-in to generate a new value. For example:
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: uid=jsmith,ou=people,dc=example,dc=com
objectClass: top
objectClass: person
objectClass: posixAccount
uid: jsmith
cn: John Smith
uidNumber: magic
....
T he DNA Plug-in only generates new, unique values. If an entry is added or modified to use a specific
value for an attribute controlled by the DNA Plug-in, the specified number is used; the DNA Plug-in will
not overwrite it.
192
Chapter 5. Managing Attributes and Values
IMPORTANT
T he DNA Plug-in only works on a single backend; it cannot manage number assignments for
multiple databases. T he DNA plug-in uses the sort control when checking whether a value has
already been manually allocated outside of the DNA Plug-in. T his validation, using the sort
control, only works on a single backend.
5.4.1. Looking at the DNA Plug-in Syntax
T he DNA Plug-in itself is a container entry, similar to the Password Storage Schemes Plug-in. Each DNA
entry underneath the DNA Plug-in entry defines a new managed range for the DNA Plug-in.
T o set new managed ranges for the DNA Plug-in, create entries beneath the container entry.
T he most basic configuration is to set up distributed numeric assignments on a single server, meaning
the ranges will not be shared or transferred between servers. A basic DNA configuration entry defines
four things:
T he attribute that's value is being managed, set in the dnaType attribute
T he entry DN to use as the base to search for entries, set in the dnaScope attribute
T he search filter to use to identify entries to manage, set in the dnaFilter attribute
T he next available value to assign, set in the dnaNextValue attribute (after the entry is created, this
is handled by the plug-in)
For example:
dn: cn=Account UIDs,cn=Distributed Numeric Assignment Plugin,cn=plugins,cn=config
objectClass: top
objectClass: extensibleObject
cn: Account UIDs
dnatype: uidNumber
dnafilter: (objectclass=posixAccount)
dnascope: ou=people,dc=example,dc=com
dnaNextValue: 1
If multiple suppliers are configured for distributed numeric assignments, then the entry must contain the
required information to transfer ranges:
T he maximum number that the server can assign; this sets the upward bound for the range, which is
logically required when multiple servers are assigning numbers. T his is set in the dnaMaxValue
attribute.
T he threshold where the range is low enough to trigger a range transfer, set in the dnaThreshold
attribute. If this is not set, the default value is 1.
A timeout period so that the server does not hang waiting for a transfer, set in the
dnaRangeRequestTimeout attribute. If this is not set, the default value is 10, meaning 10 seconds.
A configuration entry DN which is shared among all supplier servers, which stores the range
information for each supplier, set in the dnaSharedCfgDN attribute.
T he specific number range which could be assigned by the server is defined in the dnaNextRange
attribute. T his shows the next available range for transfer and is managed automatically by the plug-in,
as ranges are assigned or used by the server. T his range is just "on deck." It has not yet been
assigned to another server and is still available for its local Directory Server to use.
193
Red Hat D irectory Server 8.2 Administration Guide
dn: cn=Account UIDs,cn=Distributed Numeric Assignment Plugin,cn=plugins,cn=config
objectClass: top
objectClass: extensibleObject
cn: Account UIDs
dnatype: uidNumber
dnafilter: (objectclass=posixAccount)
dnascope: ou=People,dc=example,dc=com
dnanextvalue: 1
dnaMaxValue: 1300
dnasharedcfgdn: cn=Account UIDs,ou=Ranges,dc=example,dc=com
dnathreshold: 100
dnaRangeRequestTimeout: 60
dnaNextRange: 1301-2301
T he dnaNextRange attribute should be set explicitly only if a separate, specific range has to be
assigned to other servers. Any range set in the dnaNextRange attribute must be unique from the
available range for the other servers to avoid duplication. If there is no request from the other servers
and the server where dnaNextRange is set explicitly has reached its set dnaMaxValue, the next set of
values (part of the dnaNextRange) is allocated from this deck.
T he dnaNextRange allocation is also limited by the dnaThreshold attribute that is set in the DNA
configuration. Any range allocated to another server for dnaNextRange cannot violate the threshold for
the server, even if the range is available on the deck of dnaNextRange.
NOTE
If the dnaNextRange attribute is handled internally if it is not set explicitly. When it is handled
automatically, the dnaMaxValue attribute serves as upper limit for the next range.
All of the attributes available for a DNA Plug-in instance are listed in T able 5.3, “DNA Entry Attributes”.
194
Chapter 5. Managing Attributes and Values
T able 5.3. DNA Entry Attributes
Plug-in Attribute
Description
cn
Gives a unique name for the plug-in instance.
dnaT ype
Contains the name of the attribute for which
unique numbers are assigned.
If a prefix will be prepended to the generated
value, then be sure to use an attribute which
allows the syntax of the combined attribute value,
such as a custom attribute which allows
alphanumeric strings. Otherwise, syntax
validation will enforce the defined syntax for the
value, such as integer for uidNumber and
gidNumber, and the DNA operations will fail with
syntax violations.
dnaScope
Sets the base DN to use to search for entries to
which to apply the managed ranges.
dnaFilter
Gives an LDAP filter to use to specify the kinds of
entries for the plug-in to manage.
dnaNextValue
Gives the next available number to assign. T his
is initially set manually when the entry is created;
afterward, it is managed by the plug-in.
dnaMaxValue
Optionally, the upper limit of the range that the
server can assign. Defining the range is required
when there are multiple servers assigning
numbers to entries. T he default value is -1, which
is the same as the highest 64-bit integer.
dnaT hreshold
Sets a limit on the amount of remaining available
numbers before the server requests a new range.
dnaSharedCfgDN
Specifies the DN of a container entry that each
supplier server shares. T he plug-in automatically
creates an entry for the individual instances
underneath this entry which contains their
available ranges. T he plug-in can use this
information to request and transfer ranges as
servers consume their available range.
dnaNextRange
Shows the next range of numbers which are
available to be transferred. T his attribute can be
set automatically by the plug-in according to the
threshold and shared configuration information;
this can also be set manually for an administrator
to specifically assign an additional range of
values to a server. T his attribute is always limited
by the dnaT hreshold settings.
dnaRangeRequestT imeout
Sets a timeout period for a range request so that
a server does not hang indefinitely waiting for a
transfer.
dnaMagicRegen
Sets a word or number (outside of the assigned
range) which automatically triggers the plug-in to
assign a number to an attribute. T his is a useful
195
Red Hat D irectory Server 8.2 Administration Guide
default to use for importing entries.
dnaPrefix
Sets a string to insert in front of whatever number
is assigned. For example, if the prefix is user and
the assigned number for the attribute is 1001,
then the final assigned value is user1001.
dnaPrefix can hold any kind of string. However,
some possible values for dnaType (such as
uidNumber and gidNumber) require integer
syntax for attribute values. T o use a prefix string,
consider using a custom attribute for dnaType
which allows the syntax of the prefix plus the
generated number assignment.
5.4.2. Configuring Unique Number Assignments
T he unique number distribution is configured by creating different instances of the DNA Plug-in. T hese
DNA Plug-in instances can only be created through the command line, but they can be edited through
the Directory Server Console.
5.4 .2.1. Configuring Unique Number Assignments
NOTE
Any attribute which has a unique number assigned to it must have an equality index set for it. T he
server must perform a sorted search, internally, to see if the dnaNextvalue is already taken,
which requires an equality index on an integer attribute, with the proper ordering matching rule.
Creating indexes is described in Section 7.2, “Creating Standard Indexes”.
NOTE
Set up the DNA Plug-in on every supplier server, and be careful not to overlap the number range
values.
1. Create the shared container entry in the replicated subtree. For example:
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389
-h server.example.com
dn: ou=Ranges,dc=example,dc=coma
objectclass: top
objectclass: extensibleObject
objectclass: organizationalUnit
ou: Ranges
dn: cn=Account UIDs,ou=Ranges,dc=example,dc=coma
objectclass: top
objectclass: extensibleObject
cn: Account UIDs
196
Chapter 5. Managing Attributes and Values
2. Enable the DNA Plug-in. By default, the plug-in entry (which is the container entry) is disabled.
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=Distributed Numeric Assignment Plugin,cn=plugins,cn=config
changetype: modify
replace: nsslapd-pluginEnabled
nsslapd-pluginEnabled: on
3. Create the new DNA Plug-in instance beneath the container entry.
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389
-h server.example.com
dn: cn=Account UIDs,cn=Distributed Numeric Assignment
Plugin,cn=plugins,cn=config
objectClass: top
dnanextvalue: 1
dnaMaxValue: 1300
dnasharedcfgdn: cn=Account UIDs,ou=Ranges,dc=example,dc=com
dnathreshold: 100
dnaRangeRequestTimeout: 60
dnaMagicRegen: magic
4. Restart the server to load the new plug-in instance.
service dirsrv restart instance_name
5.4 .2.2. Editing the DNA Plug-in in the Console
NOTE
Any attribute which has a unique number assigned to it must have an equality index set for it. T he
server must perform a sorted search, internally, to see if the dnaNextvalue is already taken,
which requires an equality index on an integer attribute, with the proper ordering matching rule.
Creating indexes is described in Section 7.2, “Creating Standard Indexes”.
T he Directory Server Console can be used to edit the DNA Plug-in instances.
1. Click the Directory tab.
2. Open the config folder, and then expand the plugins folder.
3. Click the Distributed Num eric Assignm ent plug-in folder. All of the DNA Plug-in instances
are listed in the main window.
197
Red Hat D irectory Server 8.2 Administration Guide
4. Highlight the DNA instance entry, and right-click on the Advanced link to open the property editor.
5. Edit the DNA-related attributes.
198
Chapter 6. Managing the D irectory Schema
Chapter 6. Managing the Directory Schema
Red Hat Directory Server comes with a standard schema that includes hundreds of object classes and
attributes. While the standard object classes and attributes should meet most deployments'
requirements, it can be necessary to extend the schema for specific directory data. Extending the
schema is done by creating new object classes and attributes.
T he Directory Server Schema Reference is a references for most the standard Directory Server
attributes and object classes, with information on allowed and required attributes, which object classes
take which attribute, and OID and value information. T his is a good resource for identifying useful
schema elements for a directory and determining what custom schema needs to be created.
6.1. Overview of Schema
T he directory schema is a set of rules that defines how data can be stored in the directory. Directory
information is stored discrete entries, and each entry is comprised of a set of attributes and their values.
T he kind of identity being described in the entry is defined in the entry's object classes. An object class
specifies the kind of object the entry describes through the defined set of attributes for the object class.
In LDAP, an object class defines the set of attributes that can be used to define an entry. T he LDAP
standard provides object classes for many common types of entries, including people, groups, locations,
organizations and divisions, and equipment. T he identity is described in a directory entries with
attributes and their values, pairs are called attribute-value assertions or AVAs. Any piece of information
in the directory is associated with a descriptive attribute. Other aspects of the Directory Server
configuration, including matching rules and LDAP controls, are also defined in the schema. All of these
together are schema elements.
Every schema element is identified by a unique, dot-separated number. T his is called the object identifier
or OID.
6.1.1. Default Schema Files
T he schema for Directory Server is defined in several different schema files (LDIF files which define
schema elements). T he Directory Server schema files are instance-specific and are located in the
/etc/dirsrv/slapd-instance_name/schem a directory. T here is also a common
/etc/dirsrv/schem a directory; the files in this directory are used as templates for new Directory
Server instances. Putting custom schema in the /etc/dirsrv/schem a directory means that it is
automatically included with any new instances.
T he default schema files are listed and described in the Schema Reference, which also describes the
common standard attributes and object classes. T he attributes used by the Directory Server to perform
operations and manage entries is described with other configuration settings in the Configuration and
Command-Line Tool Reference.
6.1.2. Object Classes
In LDAP, an object class defines the set of attributes that can be used to define an entry. T he LDAP
standard provides object classes for many common types of entries, such as people (person and
inetOrgPerson), groups (groupOfUniqueNam es), locations (locality), organizations and
divisions (organization and organizationalUnit), and equipment (device).
In a schema file, an object class is identified by the objectclasses line, then followed by its OID, name,
a description, its direct superior object class (an object class which is required to be used in conjunction
with the object class and which shares its attributes with this object class), and the list of required
(MUST ) and allowed (MAY) attributes.
199
Red Hat D irectory Server 8.2 Administration Guide
T his is shown in Example 6.1, “person Object Class Schema Entry”.
Example 6.1. person Object Class Schema Entry
objectClasses: ( 2.5.6.6 NAME 'person' DESC 'Standard LDAP objectclass' SUP top
MUST ( sn $ cn ) MAY ( description $ seeAlso $ telephoneNumber $ userPassword )
X-ORIGIN 'RFC 2256' )
Every object class defines a number of required attributes and of allowed attributes. Required attributes
must be present in entries using the specified object class, while allowed attributes are permissible and
available for the entry to use, but are not required for the entry to be valid.
As in Example 6.1, “person Object Class Schema Entry”, the person object class requires the cn, sn,
and objectClass attributes and allows the description, seeAlso, telephoneNumber, and
userPassword attributes.
An object class can inherit attributes from another class, in addition to its own required and allowed
attributes. T he second object class is the superior or parent object class of the first.
For example, a user's entry has to have the inetOrgPerson object class. In that case, the entry must
also include the superior object class for inetOrgPerson, organizationalPerson, and the
superior object class for organizationalPerson, which is person:
objectClass:
objectClass:
objectClass:
objectClass:
top
person
organizationalPerson
inetOrgPerson
6.1.3. Attributes
Directory entries are composed of attributes and their values. T hese pairs are called attribute-value
assertions or AVAs. Any piece of information in the directory is associated with a descriptive attribute.
For instance, the cn attribute is used to store a person's full name, such as cn: John Sm ith.
Additional attributes can supply additional information about John Smith:
givenname: John
surname: Smith
mail: [email protected]
In a schema file, an attribute is identified by the attributetypes line, then followed by its OID, name, a
description, syntax (allowed format for its value), optionally whether the attribute is single- or multivalued, and where the attribute is defined.
T his is shown in Example 6.2, “description Attribute Schema Entry”.
Example 6.2. description Attribute Schema Entry
attributetypes: ( 2.5.4.13 NAME 'description' DESC 'Standard LDAP attribute type'
SYNTAX 1.3.6.1.4.1.1466.115.121.1.15 X-ORIGIN 'RFC 2256' )
200
Chapter 6. Managing the D irectory Schema
6.1.4. Extending the Schema
New, custom attributes and object classes can be added to a Directory Server instance to extend the
schema, and there are several ways to add schema elements. Using the Directory Server Console or
LDAP tools adds schema elements to the default custom schema file for an instance, 99user.ldif. It
is also possible to create a new, separate schema file and include it with the default schema files.
Adding new schema elements requires three things:
1. Planning and defining OIDs for the new schema. Schema elements are recognized by the server
by their OID, so it is important for the OIDs to be unique and organized. Directory Server itself
does not manage OIDs, but there are some best practices described in Section 6.2, “Managing
Object Identifiers”.
2. Create the new attributes. Attribute definitions require a name, a syntax (the allowed format of the
values), an OID, and a description of whether the attribute can only be used once per entry or
multiple times.
3. Create an object class to contain the new attributes. An object class lists the required attributes
for that entry type and the allowed (permissible) attributes. Because the default schema should
never be altered, if any new attributes are created, then they should be added to a custom object
class.
T he schema elements should be planned in advance; do not use multiple attributes for the same
information. Whenever possible, use the standard Directory Server schema. Directory Server has
hundreds of attributes and dozens of object classes defined in the default schema files. T he Directory
Server Schema Reference lists and describes the standard attributes and object classes; all of the
schema can be viewed in the Directory Server Console or read in the schema files in
/etc/dirsrv/slapd-instance_name/schem a. Become familiar with the available schema; then plan
what information attributes are missing and how best to fill those gaps with custom attributes. Planning
the schema is covered in the Deployment Guide.
CAUTION
T he default object classes and attributes in Directory Server are based on LDAP and X.500
standards and RFCs. Using standard schema makes the Directory Server more easily integrated
with other applications and servers and allows interoperability with LDAP clients, legacy Directory
Server instances, and future release. It is inadvisable for you to edit the standard attributes or
change the object classes.
Keep the following rules in mind when customizing the Directory Server schema:
Keep the schema as simple as possible.
Reuse existing schema elements whenever possible.
Minimize the number of mandatory attributes defined for each object class.
Do not define more than one object class or attribute for the same purpose.
Do not modify any existing definitions of attributes or object classes.
NOTE
Never delete or replace the standard schema. Doing so can lead to compatibility problems with
other directories or other LDAP client applications.
201
Red Hat D irectory Server 8.2 Administration Guide
T he schema is loaded into the Directory Server instance when the instance is started; any new schema
files are not loaded until the Directory Server is restarted or unless a reload task is initiated. Always, any
custom schema is loaded before the standard schema.
6.1.5. Schema Replication
Schema elements are replicated by default whenever the schema is changed in the Directory Server
Console or through ldapm odify because the changes are logged in the server change log. However,
replicated schema elements are all copied to 99user.ldif regardless of what schema file held the
schema on the supplier server, so replication can lead to multiple files defining the same schema.
T o keep from having duplicate schema definitions in multiple files, and for the easiest maintenance of the
schema, keep all custom schema in the default custom schema file, 99user.ldif in
/etc/dirsrv/slapd-instance_name/schem a.
When new schema files are added to the server and loaded dynamically into the Directory Server, the
schema is only loaded locally because the schema reload operation is a local task. Since the changes
do not show up in the changelog, these newly-loaded schema files are not replicated. In order to load
those schema files on any other servers, the schema has to be copied over, and then a local reload task
has to be run.
6.2. Managing Object Identifiers
Each LDAP object class or attribute must be assigned a unique name and object identifier (OID). An OID
is a dot-separated number which identifies the schema element to the server. OIDs can be hierarchical,
with a base OID that can be expanded to accommodate different branches. For example, the base OID
could be 1, and there can be a branch for attributes at 1.1 and for object classes at 1.2.
NOTE
It is not required to have a numeric OID for creating custom schema, but Red Hat strongly
recommends it for better forward compatibility and performance.
OIDs are assigned to an organization through the Internet Assigned Numbers Authority (IANA), and
Directory Server does not provide a mechanism to obtain OIDs. T o get information about obtaining OIDs,
visit the IANA website at http://www.iana.org/cgi-bin/enterprise.pl.
After obtaining a base OID from IANA, plan how the OIDs are going to be assigned to custom schema
elements. Define a branch for both attributes and object classes; there can also be branches for
matching rules and LDAP controls.
Once the OID branches are defined, create an OID registry to track OID assignments. An OID registry is
a list that gives the OIDs and descriptions of the OIDs used in the directory schema. T his ensures that
no OID is ever used for more than one purpose. Publish the OID registry with the custom schema.
6.3. Directory Server Attribute Syntaxes
T he attribute's syntax defines the format of the values which the attribute allows; as with other schema
elements, the syntax is defined for an attribute using the syntax's OID in the schema file entry. In the
Directory Server Console, the syntax is referenced by its friendly name.
T he Directory Server uses the attribute's syntax to perform sorting and pattern matching on entries.
202
Chapter 6. Managing the D irectory Schema
For more information about LDAP attribute syntaxes, see RFC 4517.
203
Red Hat D irectory Server 8.2 Administration Guide
T able 6.1. Supported LDAP Attribute Syntaxes
Name
OID
Definition
Binary
1.3.6.1.4.1.1466.115.121.1.5
For values which are binary.
Bitstring
1.3.6.1.4.1.1466.115.121.1.6
For values which are bitstings.
T his is supported for searches
which require accessing
attributes using bitstring flags.
Boolean
1.3.6.1.4.1.1466.115.121.1.7
For attributes with only two
allowed values, true or false.
Country String
1.3.6.1.4.1.1466.115.121.1.11
For values which are limited to
exactly two printable string
characters; for example, US for
the United States.
DN
1.3.6.1.4.1.1466.115.121.1.12
For values which are DNs.
Delivery Method
1.3.6.1.4.1.1466.115.121.1.14
For values which are contained
a preferred method of delivering
information or contacting an
entity. T he different values are
separated by a dollar sign ($).
For example:
telephone $ physical
DirectoryString (CaseInsensitive String)
1.3.6.1.4.1.1466.115.121.1.15
For values which are caseinsensitive strings.
Enhanced Guide
1.3.6.1.4.1.1466.115.121.1.21
For values which contain
complex search parameters
based on attributes and filters.
Facsimile
1.3.6.1.4.1.1466.115.121.1.22
For values which contain fax
numbers.
Fax
1.3.6.1.4.1.1466.115.121.1.23
For values which contain the
images of transmitted faxes.
GeneralizedT ime
1.3.6.1.4.1.1466.115.121.1.24
For values which are encoded
as printable strings. T he time
zone must be specified. It is
strongly recommended to use
GMT time.
Guide
1.3.6.1.4.1.1466.115.121.1.25
Obsolete. For values which
contain complex search
parameters based on attributes
and filters.
IA5String (Case-Exact String)
1.3.6.1.4.1.1466.115.121.1.26
For values which are case-exact
strings.
Integer
1.3.6.1.4.1.1466.115.121.1.27
For values which are whole
numbers.
JPEG
1.3.6.1.4.1.1466.115.121.1.28
For values which contain image
data.
Name and Optional UID
1.3.6.1.4.1.1466.115.121.1.34
For values which contain a
combination value of a DN and
204
Chapter 6. Managing the D irectory Schema
(optional) unique ID.
Numeric String
1.3.6.1.4.1.1466.115.121.1.36
For values which contain a
string of both numerals and
spaces.
OctetString
1.3.6.1.4.1.1466.115.121.1.40
For values which are binary; this
is the same as using the binary
syntax.
OID
1.3.6.1.4.1.1466.115.121.1.37
For values which contain an
object identifier (OID).
Postal Address
1.3.6.1.4.1.1466.115.121.1.41
For values which are encoded in
the format postal-address =
dstring * ("$" dstring). For
example:
1234 Main St.$Raleigh,
NC 12345$USA
Each dstring component is
encoded as a DirectoryString
value. Backslashes and dollar
characters, if they occur, are
quoted, so that they will not be
mistaken for line delimiters.
Many servers limit the postal
address to 6 lines of up to thirty
characters.
PrintableString
1.3.6.1.4.1.1466.115.121.1.58
For values which contain strings
containing alphabetic, numeral,
and select punctuation
characters (as defined in RFC
4517).
Space-Insensitive String
2.16.840.1.113730.3.7.1
For values which contain spaceinsensitive strings.
T elephoneNumber
1.3.6.1.4.1.1466.115.121.1.50
For values which are in the form
of telephone numbers. It is
recommended to use telephone
numbers in international form.
T eletex T erminal Identifier
1.3.6.1.4.1.1466.115.121.1.51
For values which contain an
international telephone number.
T elex Number
1.3.6.1.4.1.1466.115.121.1.52
For values which contain a telex
number, country code, and
answerback code of a telex
terminal.
URI
For values in the form of a URL,
introduced by a string such as
http://, https://, ftp://,
ldap://, and ldaps://. T he
URI has the same behavior as
IA5String. See RFC 4517 for
more information on this syntax.
205
Red Hat D irectory Server 8.2 Administration Guide
6.4. Managing Custom Schema in the Console
T he Directory Server Console shows all attributes in the schema, and custom attributes can be created,
edited, and deleted from the schema.
Section 6.4.1, “Viewing Attributes and Object Classes”
Section 6.4.2, “Creating Attributes”
Section 6.4.3, “Creating Object Classes”
Section 6.4.4, “Editing Custom Schema Elements”
Section 6.4.5, “Deleting Schema”
6.4.1. Viewing Attributes and Object Classes
All of the information about the attributes and object classes which are currently loaded in the server
instance are visible with the other server configuration.
1. In the Directory Server Console, select the Configuration tab.
2. In the left navigation tree, select the Schem a folder.
3. T here are three tabs which display the schema elements loaded in Directory Server: Object
Class, Attributes, and Matching Rules.
T he Attributes tab is broken into two sections for default and custom attributes. Both sections show
the attribute name, OID, syntax, and whether the attribute is multi-valued.
206
Chapter 6. Managing the D irectory Schema
T he Object Classes tab shows the list of object classes on the left. When an object class is
highlighted, its OID and superior object class are listed in the fields at the top and its required and
allowed attributes are listed in the boxes on the right.
207
Red Hat D irectory Server 8.2 Administration Guide
6.4.2. Creating Attributes
NOTE
After adding new attributes to the schema, create a new object class to contain them, as
described in Section 6.4.3, “Creating Object Classes”.
1. Select the Configuration tab.
2. In the left navigation tree, select the Schem a folder, and then select the Attributes tab in the
right pane.
3. Click Create.
4. Fill in the information for the new attribute.
T he attribute name; this must be unique.
T he OID; this is not required, but for compatibility and server performance, assigning a unique
numeric OID is strongly recommended.
208
Chapter 6. Managing the D irectory Schema
T he syntax; this is the allowed format for the attributes values.
Whether the attribute is multi-valued; by default, all attributes can be used more than once in
an entry, but deselecting the checkbox means the attribute can be used only once.
5. Click OK.
6.4.3. Creating Object Classes
A new object class must be created with a unique name, a parent object, and required and optional
attributes. T o create an object class:
1. In the Directory Server Console, select the Configuration tab.
2. In the navigation tree, select the Schem a folder, and then select the Object Classes tab in the
right pane.
3. Click the Create button in the Object Classes tab.
4. Fill in the information about the new object class.
209
Red Hat D irectory Server 8.2 Administration Guide
T he name; this must be unique.
T he OID; this is not required, but for compatibility and server performance, assigning a unique
numeric OID is strongly recommended.
T he superior object class for the entry. T he default is top; selecting another object class
means that the new object class inherits all of the required and allowed attributes from the
parent, in addition to its own defined attributes.
Required and allowed attributes. Select the attributes on the left and used the Add buttons by
the Available Attributes and Required Attributes boxes to add the attributes as
appropriate.
NOTE
Attributes that are inherited from the parent object classes cannot be removed, regardless
of whether they are allowed or required.
5. Click OK to save the new object class.
6.4.4. Editing Custom Schema Elements
Only user-created attributes or object classes can be edited; standard schema elements cannot be
edited.
1. In the Directory Server Console, select the Configuration tab.
2. In the left navigation tree, select the Schem a folder.
210
Chapter 6. Managing the D irectory Schema
3. Open the Object Classes or Attributes tab.
4. Select the schema element to edit from the list. Only custom (user-defined) schema can be edited
in the Directory Server Console.
5. Click the Edit button at the bottom of the window.
6. Edit any of the schema information.
6.4.5. Deleting Schema
Only user-created attributes or object classes can be deleted; standard schema elements cannot be
deleted.
1. In the Directory Server Console, select the Configuration tab.
2. In the left navigation tree, select the Schem a folder.
3. Open the Object Classes or Attributes tab.
4. Select the schema element to delete from the list. Only custom (user-defined) schema can be
deleted in the Directory Server Console.
5. Click the Delete button at the bottom of the window.
6. Confirm the deletion.
211
Red Hat D irectory Server 8.2 Administration Guide
WARNING
T he server immediately deletes the schema element. T here is no undo.
6.5. Managing Schema Using ldapmodify
As with the Directory Server Console, ldapm odify can be used to add, edit, and delete custom
schema elements. ldapm odify also modifies the default custom schema file for a Directory Server
instance, 99user.ldif.
6.5.1. Creating Attributes
A custom attribute entry is itself an attributetypes entry for the cn=schem a entry. T he
attributetypes attribute has the format:
attributetypes: ( definition )
T he definition contains five components:
An OID, usually a dot-separated number
A unique name, in the form NAME name
A description, in the form DESC description
T he OID for the syntax of the attribute values, listed in T able 6.1, “Supported LDAP Attribute
Syntaxes”, in the form SYNT AX OID
Optionally, the source where the attribute is defined
T he attribute definition is added to the custom schema file, 99user.ldif, by by running an LDAP
command and modifying the cn=schem a entry. For example:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -v
dn: cn=schema
changetype: modify
add: attributetypes
attributetypes: ( 1.2.3.4.5.6.1 NAME 'dateofbirth' DESC 'For employee birthdays'
SYNTAX 1.3.6.1.4.1.1466.115.121.1.15 SINGLE-VALUED X-ORIGIN 'Example defined')
6.5.2. Creating Object Classes
An object class definition is an objectclasses attribute for the cn=schem a entry. T he objectclasses
attribute has the format:
objectclasses: ( definition )
T he object class definition contains several components:
An OID, usually a dot-separated number
A unique name, in the form NAME name
A description, in the form DESC description
T he superior, or parent, object class for this object class, in the form SUP object_class; if there is no
related parent, use SUP top
212
Chapter 6. Managing the D irectory Schema
related parent, use SUP top
T he word AUXILIARY, which gives the type of entry to which the object class applies; AUXILIARY
means it can apply to any entry
A list of required attributes, preceded by the word MUST ; to include multiple attributes, enclose the
group in parentheses and separate with attributes with dollar signs ($)
A list of allowed attributes, preceded by the word MAY; to include multiple attributes, enclose the
group in parentheses and separate with attributes with dollar signs ($)
T he object class definition is added to the custom schema file, 99user.ldif, by by running an LDAP
command and modifying the cn=schem a entry. For example:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com -v
dn: cn=schema
changetype: modify
add: objectclasses
objectclasses: ( 2.16.840.1133730.2.123 NAME 'examplePerson' DESC 'Example Person
Object Class' SUP inetOrgPerson AUXILIARY MUST cn MAY (exampleDateOfBirth $
examplePreferredOS) )
6.5.3. Deleting Schema
WARNING
Never delete default schema elements. T hose are required by the Directory Server to run.
1. Remove the unwanted attributes from any entries which use them, then from any object classes in
the schema file which accept that attribute. Likewise, to remove an object class, remove it from any
entries.
2. Run ldapm odify to remove the attribute. For example:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=schema
changetype: modify
delete: objectclasses
objectclasses: ( 2.16.840.1133730.2.123 NAME 'examplePerson' DESC 'Example
Person Object Class' SUP inetOrgPerson AUXILIARY MUST cn MAY
(exampleDateOfBirth $ examplePreferredOS) )
CAUTION
Be sure to specify the exact object class or attribute to remove; using only the
attributetypes or objectclasses attribute without the value will delete every userdefined attribute or object class in the file.
If the custom attribute or object class is in a custom schema file other than 99user.ldif, edit the file
directly. Neither the Directory Server Console nor LDAP tools can edit a schema file other than
99user.ldif.
213
Red Hat D irectory Server 8.2 Administration Guide
6.6. Creating Custom Schema Files
Schema files are simple LDIF files which define the cn=schem a entry. Each attribute and object class is
added as an attribute to that entry. Here are the requirements for creating a schema file:
T he first line must be dn: cn=schem a.
T he schema file can include both attributes and object classes, but it can also include only one or the
other.
If both attributes and object classes are defined in the style, all of the attributes must be listed in the
file first, then all of the attributes must be listed first, then the object classes.
T he object classes can use attributes defined in other schema files.
T he file must be named in the format [1-9][0-9]text.ldif.
T he file must always begin with two numbers. Numerically, the schema file cannot be loaded after the
core configuration schema (which begin with 00 and 01).
Also, the Directory Server always writes its custom schema to the numerically and alphabetically
highest named schema file in the schema directory. It expects this file to be 99user.ldif. If this file
is not 99user.ldif, the server can experience problems. So, always make sure custom schema
files are at least alphabetically lower than 99user.ldif. T he name 99alpha.ldif is okay; the
name 99zzz.ldif is not.
Practices for creating schema files are described in more detail in the Deployment Guide.
Attributes are defined in the schema file as attributetypes attributes to the schema, with five
components:
An OID, usually a dot-separated number
A unique name, in the form NAME name
A description, in the form DESC description
T he OID for the syntax of the attribute values, listed in T able 6.1, “Supported LDAP Attribute
Syntaxes”, in the form SYNT AX OID
Optionally, the source where the attribute is defined
For example:
attributetypes: ( 1.2.3.4.5.6.1 NAME 'dateofbirth' DESC 'For employee birthdays'
SYNTAX 1.3.6.1.4.1.1466.115.121.1.15 SINGLE-VALUED X-ORIGIN 'Example defined')
Likewise, object classes are defined as objectclasses attributes, although there is slightly more
flexibility in how the object class is defined. T he only required configurations are the name and OID for
the object class; all other configuration depends on the needs for the object class:
An OID, usually a dot-separated number
A unique name, in the form NAME name
A description, in the form DESC description
T he superior, or parent, object class for this object class, in the form SUP object_class; if there is no
related parent, use SUP top
T he word AUXILIARY, which gives the type of entry to which the object class applies; AUXILIARY
means it can apply to any entry
A list of required attributes, preceded by the word MUST ; to include multiple attributes, enclose the
group in parentheses and separate with attributes with dollar signs ($)
214
Chapter 6. Managing the D irectory Schema
A list of allowed attributes, preceded by the word MAY; to include multiple attributes, enclose the
group in parentheses and separate with attributes with dollar signs ($)
For example:
objectclasses: ( 2.16.840.1133730.2.123 NAME 'examplePerson' DESC 'Example Person
Object Class' SUP inetOrgPerson AUXILIARY MUST cn MAY (exampleDateOfBirth $
examplePreferredOS) )
Example 6.3, “Example Schema File” shows a simplified schema file.
Example 6.3. Example Schema File
dn: cn=schema
attributetypes: ( 2.16.840.1133730.1.123 NAME 'dateofbirth' DESC 'For employee
birthdays' SYNTAX 1.3.6.1.4.1.1466.115.121.1.15 X-ORIGIN 'Example defined')
objectclasses: ( 2.16.840.1133730.2.123 NAME 'examplePerson' DESC 'Example
Person Object Class' SUP inetOrgPerson AUXILIARY MAY (dateofbirth) )
Custom schema files should be added to the Directory Server instance's schema directory,
/etc/dirsrv/slapd-instance_name/schem a. T he schema in these files are not loaded and
available to the server unless the server is restarted or a dynamic reload task is run.
6.7. Dynamically Reloading Schema
By default, the schema files used by the Directory Server instance are loaded into the directory when it is
started. T his means that any new schema files added to the schema directory are not available for use
unless the server is restarted. T he Directory Server has a task which manually reloads the full schema
for the Directory Server instance, including custom files, without requiring a server restart.
T he schema reload task can be initiated in two ways:
Using the schem a-reload.pl script
Adding a cn=schem a reload task entry using ldapm odify
6.7.1. Reloading Schema Using schema-reload.pl
T he schem a-reload.pl script launches a special task to reload all of the schema files used by a
specific Directory Server instance. T his allows custom schema files to be loaded dynamically without
having to add schema elements to 99user.ldif.
1. Open the tool directory for the Directory Server instance,
/usr/lib/dirsrv/slapd-instance_name/.
2. Run the script, binding as the Directory Manager.
./schema-reload.pl -D "cn=Directory Manager" -w secret
T he Directory Server responds that it has added the new reload task entry.
adding new entry cn=schema_reload_2009_1_6_17_52_4,cn=schema reload
task,cn=tasks,cn=config
T his reloads the schema from the default schema directory,
215
Red Hat D irectory Server 8.2 Administration Guide
/etc/dirsrv/slapd-instance_name/schem a, which is recommended. It is also possible to
specify a different directory using the -d option.
./schema-reload.pl -D "cn=Directory Manager" -w password -d /export/customschema
IMPORTANT
All of the schema is reloaded with the schema reload task, not just the newest schema or
modified schema. T his means that whatever directory is specified should contain the full
schema for the Directory Server, or the Directory Server instance may have serious
performance problems.
T he schem a-reload.pl is described in more detail in the Configuration and Command-Line Tool
Reference.
6.7.2. Reloading Schema Using ldapmodify
T he schem a-reload.pl script creates a special task entry in a Directory Server instance which
reloads schema files; it is also possible to reload schema by creating the task entry directly. T ask
entries occur under the cn=tasks configuration entry in the dse.ldif file, so it is also possible to
initiate a task by adding the entry using ldapm odify. As soon as the task is complete, the entry is
removed from the directory.
T o initiate a schema reload task, add an entry under the cn=schem a reload
task,cn=tasks,cn=config entry. T he only required attribute is the cn for the specific task.
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=example schema reload,cn=schema reload task,cn=tasks,cn=config
objectclass: extensibleObject
cn:example schema reload
T he default schema directory from which the Directory Server instance reloads the schema is in
/etc/dirsrv/slapd-instance_name/schem a; it is possible to specify a different schema directory
using the schemadir attribute, which is analogous to the -d option with schem a-reload.pl.
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=example schema reload,cn=schema reload task,cn=tasks,cn=config
objectclass: extensibleObject
cn:example schema reload
schemadir: /export/schema
216
Chapter 6. Managing the D irectory Schema
IMPORTANT
All of the schema is reloaded with the schema reload task, not just the newest schema or
modified schema. T his means that whatever directory is specified should contain the full schema
for the Directory Server, or the Directory Server instance may have serious performance
problems.
As soon as the task is completed, the entry is deleted from the dse.ldif configuration, so it is possible
to reuse the same task entry continually.
T he cn=schem a reload task configuration is described in more detail in the Configuration and
Command-Line Tool Reference.
6.7.3. Reloading Schema with Replication
T he schema reload task is a local operation, so schema changes are not replicated in a multi-master
environment if the schema is added to one supplier but not to the others. T o load the new schema files
on all of the supplier servers:
1. Stop replication.
2. Copy the new schema file over and run the schema reload task for every supplier and replica
server.
3. Restart replication.
6.7.4. Schema Reload Errors
When the schema reload task runs, the command prompt only shows that the task is initiated.
adding new entry cn=schema reload task 1, cn=schema reload task, cn=tasks,
cn=config
However, the task does not return whether it completed successfully. T o verify the schema reload
operation was successful, check the error logs. T he schema reload has two tasks, first validating the
schema file and then loading it.
A success message shows that the validation passed and the task finished.
[06/Jan/2009:17:52:04 -0500] schemareload - Schema reload task starts (schema dir:
default) ...
[06/Jan/2009:17:52:04 -0500] schemareload - Schema validation passed.
[06/Jan/2009:17:52:04 -0500] schemareload - Schema reload task finished.
If there is a failure, then the logs show which step failed and why.
[..]
[..]
[..]
[..]
schemareload - Schema reload task starts (schema dir: /bogus) ...
schema - No schema files were found in the directory /bogus
schema_reload - schema file validation failed
schemareload - Schema validation failed.
6.8. Turning Schema Checking On and Off
When schema checking is on, the Directory Server ensures three things:
217
Red Hat D irectory Server 8.2 Administration Guide
T he object classes and attributes using are defined in the directory schema.
T he attributes required for an object class are contained in the entry.
Only attributes allowed by the object class are contained in the entry.
Schema checking is turned on by default in the Directory Server, and the Directory Server should always
run with schema checking turned on. T he only situation where is may be beneficial to turn schema
checking off is to accelerate LDAP import operations. However, there is a risk of importing entries that do
not conform to the schema. Consequently, it is impossible to search for these entries.
1. In the Directory Server Console, select the Configuration tab.
2. Highlight the server icon at the top of the navigation tree, then select the Settings tab in the
right pane.
3. T o enable schema checking, check the Enable Schem a Checking checkbox; clear it to turn
off schema checking.
4. Click Save.
T o turn schema checking on and off using LDAP commands, edit the value of the nsslapdschemacheck attribute. For example:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=config
changetype: modify
replace: nsslapd-schemacheck: on
nsslapd-schemacheck: off
For information, see the Directory Server Configuration and Command-Line Tool Reference.
218
Chapter 6. Managing the D irectory Schema
6.9. Using Syntax Validation
With syntax validation, the Directory Server checks that the value of an attribute follows the rules of the
syntax given in the definition for that attribute. For example, syntax validation will confirm that a new
telephoneNumber attribute actually has a valid telephone number for its value.
6.9.1. About Syntax Validation
As with schema checking, validation reviews any directory modification and rejects changes that violate
the syntax. Additional settings can be optionally configured so that syntax validation can log warning
messages about syntax violations and then either reject the modification or allow the modification
process to succeed.
T his feature validates all attribute syntaxes, with the exception of binary syntaxes (which cannot be
verified) and non-standard syntaxes, which do not have a defined required format. T he syntaxes are
validated against RFC 4514.
6.9.2. Syntax Validation and Other Directory Server Operations
Syntax validation is mainly relevant for standard LDAP operations like creating entries (add) or editing
attributes (modify). Validating attribute syntax can impact other Directory Server operations, however.
Database Encryption
For normal LDAP operations, an attribute is encrypted just before the value is written to the database.
T his means T hat encryption occurs after the attribute syntax is validated.
Encrypted databases (as described in Section 2.2.3, “Configuring Attribute Encryption”) can be exported
and imported. Normally, it is strongly recommended that these export and import operations are done
with the -E flag with db2ldif and ldif2db, which allows syntax validation to occur just fine for the
import operation. However, if the encrypted database is exported without using the -E flag (which is not
supported), then an LDIF with encrypted values is created. When this LDIF is then imported, the
encrypted attributes cannot be validated, a warning is logged, and attribute validation is skipped in the
imported entry.
Synchronization
T here may be differences in the allowed or enforced syntaxes for attributes in Windows Active Directory
entries and Red Hat Directory Server entries. In that case, the Active Directory values could not be
properly synced over because syntax validation enforces the RFC standards in the Directory Server
entries.
Replication
If the Directory Server 8.2 instance is a supplier which replicates its changes to a consumer, then there
is no issue with using syntax validation. However, if the supplier in replication is an older version of
Directory Server or has syntax validation disabled, then syntax validation should not be used on the 8.2
consumer because the Directory Server 8.2 consumer may reject attribute values that the master allows.
6.9.3. Enabling or Disabling Syntax Validation
Syntax validation is configured by the nsslapd-syntaxcheck attribute. T he value of this attribute is
either on or off (by default, this is on). T o change the syntax validation, modify this attribute using
ldapm odify or by editing the dse.ldif file directly.
219
Red Hat D irectory Server 8.2 Administration Guide
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389
dn: cn=config
changetype: modify
replace: nsslapd-syntaxcheck
nsslapd-syntaxcheck: off
TIP
If syntax validation is disabled, then run the syntax-validate.pl script to audit existing
attribute values before re-enabling syntax validation. See Section 6.9.6, “Validating the Syntax of
Existing Attribute Values”.
6.9.4. Enabling Strict Syntax Validation for DNs
When syntax validation is enabled, DNs are validated against RFC 4514, as are other attribte syntaxes.
However, DN syntax validation is enabled separately because the strictness of later standards can
invalidate old-style DNs, and therefore directory trees.
Syntax validation checks DNs against section 3 in RFC 4514.
T he value of this attribute is either on or off (by default, this is off). T o change the syntax validation,
modify this attribute using ldapm odify or by editing the dse.ldif file directly.
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389
dn: cn=config
changetype: modify
replace: nsslapd-dn-validate-strict
nsslapd-dn-validate-strict: on
If strict DN validation is enabled and a DN value does not conform to the required syntax, then the
operation fails with LDAP result code 34, INVALID_DN_SYNT AX.
TIP
If a DN contains a nested DN, then the syntax validation checks not only the DN, but it looks up
and normalizes the nested DN as well. T his can affect performance in some situations, because
a single validation can result in multiple lookup and normalization operations.
T o disable checking on the nested DNs, turn off the nsslapd-normalize-nested-dn attribute.
T he primary DN will be validated and normalized as expected.
6.9.5. Enabling Syntax Validation Warnings (Logging)
By default, syntax validation rejects any add or modify operations where an attribute value violates the
required syntax. However, the violation itself is not recorded to the errors log by default. T he nsslapdsyntaxlogging attribute enables error logging for any syntax violations.
220
Chapter 6. Managing the D irectory Schema
NOTE
Syntax violations discovered by the syntax validation script and task are logged in the Directory
Server error log.
If nsslapd-syntaxlogging and nsslapd-syntaxcheck are both enabled, then any invalid attribute
modification is rejected and the message written to the log. If nsslapd-syntaxlogging is enabled but
nsslapd-syntaxcheck is disabled, then the operation is allowed to succeed, but the warning message
is still written to the error log.
T he value of this attribute is either on or off (by default, this is off). T o enable syntax validation logging,
edit the attribute using ldapm odify or by editing the dse.ldif file directly.
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389
dn: cn=config
changetype: modify
replace: nsslapd-syntaxlogging
nsslapd-syntaxlogging: on
6.9.6. Validating the Syntax of Existing Attribute Values
Syntax validation checks every modification to attributes to make sure that the new value has the
required syntax for that attribute type. However, syntax validation only audits changes to attribute values,
such as when an attribute is added or modified. It does not validate the syntax of existing attribute
values.
Validation of existing attribute values can be done with the syntax validation script. T his script checks
entries under a specified subtree (in the -b option) and, optionally, only entries which match a specified
filter (in the -f option). For example:
/usr/lib64/dirsrv/instance_name/syntax-validate.pl -D "cn=directory manager" -w
secret -b "ou=people,dc=example,dc=com" -f "objectclass=inetorgperson)"
Any syntax violations in existing attribute values can be identified and corrected through this script.
TIP
If syntax validation is disabled or if a server is migrated, then there may be data in the server
which does not conform to attribute syntax requirements. T he syntax validation script can be run
to evaluate those existing attribute values before enabling syntax validation.
Alternately, a task can be launched to initiate syntax validation, specifying the required base DN and,
optionally, LDAP search filter.
221
Red Hat D irectory Server 8.2 Administration Guide
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=example, cn=syntax validation,cn=tasks,cn=config
objectclass: extensibleObject
cn:example
basedn: ou=people,dc=example,dc=com
filter: "(objectclass=inetorgperson)"
222
Chapter 7. Managing Indexes
Chapter 7. Managing Indexes
Indexing makes searching for and retrieving information easier by classifying and organizing attributes
or values. T his chapter describes the searching algorithm itself, placing indexing mechanisms in context,
and then describes how to create, delete, and manage indexes.
7.1. About Indexes
T his section provides an overview of indexing in Directory Server. It contains the following topics:
Section 7.1.1, “About Index T ypes”
Section 7.1.2, “About Default, System, and Standard Indexes”
Section 7.1.3, “Overview of the Searching Algorithm”
Section 7.1.5, “Indexing Performance”
Section 7.1.6, “Balancing the Benefits of Indexing”
7.1.1. About Index Types
Indexes are stored in files in the directory's databases. T he names of the files are based on the indexed
attribute, not the type of index contained in the file. Each index file may contain multiple types of indexes
if multiple indexes are maintained for the specific attribute. For example, all indexes maintained for the
common name attribute are contained in the cn.db4 file.
Directory Server supports the following types of index:
Presence index (pres) contains a list of the entries that contain a particular attribute, which is very
useful for searched. For example, it makes it easy to examine any entries that contain access control
information. Generating an aci.db4 file that includes a presence index efficiently performs the
search for ACI=* to generate the access control list for the server.
T he presence index is not used for base object searches.
Equality index (eq) improves searches for entries containing a specific attribute value. For example,
an equality index on the cn attribute allows a user to perform the search for cn=Babs Jensen far
more efficiently.
Approximate index (approx) is used for efficient approximate or sounds-like searches. For example,
an entry may include the attribute value cn=Robert E Lee. An approximate search would return
this value for searches against cn~=Robert Lee, cn~=Robert, or cn~=Lee. Similarly, a search
against l~=San Fransisco (note the misspelling) would return entries including l=San
Francisco.
Substring index (sub) is a costly index to maintain, but it allows efficient searching against substrings
within entries. Substring indexes are limited to a minimum of three characters for each entry.
For example, searches of the form cn=* derson , match the common names containing strings
such as Bill Anderson, Jill Henderson, or Steve Sanderson. Similarly, the search for
telephoneNum ber= * 555* returns all the entries in the directory with telephone numbers that
contain 555.
International index speeds up searches for information in international directories. T he process for
creating an international index is similar to the process for creating regular indexes, except that it
applies a matching rule by associating an object identifier (OID) with the attributes to be indexed.
T he supported locales and their associated OIDs are listed in Appendix C, Internationalization. If
there is a need to configure the Directory Server to accept additional matching rules, contact Red Hat
Professional Services.
Browsing index, or virtual list view (VLV) index, speeds up the display of entries in the Directory
223
Red Hat D irectory Server 8.2 Administration Guide
Server Console. T his index is particularly useful if a branch of your directory contains hundreds of
entries; for example, the ou=people branch. You can create a browsing index on any branch point
in the directory tree to improve display performance through the Directory Server Console or by using
the vlvindex command-line tool, which is explained in the Directory Server Configuration and
Command-Line Tool Reference.
7.1.2. About Default, System, and Standard Indexes
When you install Directory Server, a set of default and system indexes is created per database instance.
T o maintain these indexes, the directory uses standard indexes.
7.1.2.1. Overview of Default Indexes
T he default indexes can be modified depending on the directory indexing needs. Always ensure that no
server plug-ins or other servers depend on a default index before removing it.
T able 7.1, “Default Indexes” lists the default indexes installed with the directory.
224
Chapter 7. Managing Indexes
T able 7.1. Default Indexes
Attribute
Eq
Pres
Sub
Purpose
cn
Improves the
performance of
the most common
types of user
directory
searches.
givenname
Improves the
performance of
the most common
types of user
directory
searches.
mail
Improves the
performance of
the most common
types of user
directory
searches.
mailHost
Used by a
messaging server.
member
Improves
Directory Server
performance. T his
index is also used
by the Referential
Integrity Plug-in.
See Section 3.5,
“Maintaining
Referential
Integrity” for more
information.
owner
Improves
Directory Server
performance. T his
index is also used
by the Referential
Integrity Plug-in.
See Section 3.5,
“Maintaining
Referential
Integrity” for more
information.
see Also
Improves
Directory Server
performance. T his
index is also used
by the Referential
Integrity Plug-in.
See Section 3.5,
225
Red Hat D irectory Server 8.2 Administration Guide
“Maintaining
Referential
Integrity” for more
information.
sn
Improves the
performance of
the most common
types of user
directory
searches.
telephoneNumber
Improves the
performance of
the most common
types of user
directory
searches.
uid
Improves
Directory Server
performance.
unique member
Improves
Directory Server
performance. T his
index is also used
by the Referential
Integrity Plug-in.
See Section 3.5,
“Maintaining
Referential
Integrity” for more
information.
7.1.2.2. Overview of System Indexes
System indexes cannot be deleted or modified. T hey are required by the directory to function properly.
T able 7.2, “System Indexes” lists the system indexes included with the directory.
226
Chapter 7. Managing Indexes
T able 7.2. System Indexes
Attribute
Eq
Pres
Purpose
aci
Allows the Directory
Server to quickly obtain
the access control
information maintained
in the database.
objectClass
Used to help accelerate
subtree searches in
the directory.
entryDN
Speeds up entry
retrieval based on DN
searches.
parentID
Enhances directory
performance during
one-level searches.
numSubordinates
Used by the Directory
Server Console to
enhance display
performance on the
Directory tab.
nsUniqueID
Used to search for
specific entries.
7.1.2.3. Overview of Standard Indexes
Because of the need to maintain default indexes and other internal indexing mechanisms, the Directory
Server also maintains certain standard index files. T he standard index, id2entry.db4 , exists by
default in Directory Server; you do not need to generate it.
T he id2entry.db4 contains the actual directory database entries. All other database files can be
recreated from this one.
7.1.3. Overview of the Searching Algorithm
Indexes are used to speed up searches. T o understand how the directory uses indexes, it helps to
understand the searching algorithm. Each index contains a list of attributes (such as the cn, common
name, attribute) and a pointer to the entries corresponding to each value. Directory Serverprocesses a
search request as follows:
1. An LDAP client application sends a search request to the directory.
2. T he directory examines the incoming request to make sure that the specified base DN matches a
suffix contained by one or more of its databases or database links.
If they do match, the directory processes the request.
If they do not match, the directory returns an error to the client indicating that the suffix does
not match. If a referral has been specified in the nsslapd-referral attribute under
cn=config, the directory also returns the LDAP URL where the client can attempt to pursue
the request.
If the search request for each database attribute can be satisfied by a single index, then the
server reads that index to generate a list of potential matches.
227
Red Hat D irectory Server 8.2 Administration Guide
If there is no index for the attribute, the directory generates a candidate list that includes all
entries in the database, which makes the search considerably slower.
If a search request contains multiple attributes, the directory consults multiple indexes and
then combines the resulting lists of candidate entries.
If there is an index for the attribute, the directory takes the candidate matches from the index
files in the form of a series of entry ID numbers.
3. T he directory uses the returned entry ID numbers to read the corresponding entries from the
id2entry.db4 file. T he Directory Server then examines each of the candidate entries to see if
any match the search criteria. T he directory returns matching entries to the client as each is
found.
T he directory continues until either it has examined all candidate entries or it reaches the limit set
in one of the following attributes:
nsslapd-sizelim it which specifies the maximum number of entries to return from a search
operation. If this limit is reached, the directory returns any entries it has located that match the
search request, as well as an exceeded size limit error.
nsslapd-tim elim it which specifies the maximum number of seconds allocated for a
search request. If this limit is reached, the directory returns any entries it has located that
match the search request, as well as an exceeded time limit error.
nsslapd-lookthroughlim it, which specifies the maximum number of entries that the
directory will check when examining candidate entries in response to a search request.
nsslapd-idlistscanlim it which specifies the maximum number of entries in an ID list
before the list is considered to equal the entire database.
See Directory Server Configuration and Command-Line Tool Reference for further information
about these attributes.
7.1.4. Approximate Searches
In addition, the directory uses a variation of the metaphone phonetic algorithm to perform searches on
an approximate index. Each value is treated as a sequence of words, and a phonetic code is generated
for each word.
NOTE
T he metaphone phonetic algorithm in Directory Server supports only US-ASCII letters. T herefore,
use approximate indexing only with English values.
Values entered on an approximate search are similarly translated into a sequence of phonetic codes. An
entry is considered to match a query if both of the following are true:
All of the query string codes match the codes generated in the entry string.
All of the query string codes are in the same order as the entry string codes.
Name in the Directory
(Phonetic Code)
Query String (Phonetic
code)
Match Comments
Alice B Sarette (ALS B SRT )
Alice Sarette (ALS SRT )
Matches. Codes are specified in
the correct order.
Alice Sarrette (ALS SRT )
Matches. Codes are specified in
the correct order, despite the
misspelling of Sarette.
228
Chapter 7. Managing Indexes
Surette (SRT )
Matches. T he generated code
exists in the original name,
despite the misspelling of
Sarette.
Bertha Sarette (BR0 SRT )
No match. T he code BR0 does
not exist in the original name.
Sarette, Alice (SRT ALS)
No match. T he codes are not
specified in the correct order.
7.1.5. Indexing Performance
Each index that the directory uses is composed of a table of index keys and matching entry ID lists. T his
entry ID list is used by the directory to build a list of candidate entries that may match a client
application's search request; Section 7.1, “About Indexes” describes each kind of Directory Server index.
T he Directory Server secondary index structure greatly improves write and search operations.
7.1.5.1. Indexing Performance
While achieving extremely high read performance, in previous versions of Directory Server, write
performance was limited by the number of bytes per second that could be written into the storage
manager's transaction log file. Large log files were generated for each LDAP write operation; in fact, log
file verbosity could easily be 100 times the corresponding number of bytes changed in the Directory
Server. T he majority of the contents in the log files are related to index changes (ID insert and delete
operations).
T he secondary index structure was separated into two levels in the old design:
T he ID list structures, which were the province of the Directory Server backend and opaque to the
storage manager.
T he storage manager structures (Btrees), which were opaque to the Directory Server backend code.
Because it had no insight into the internal structure of the ID lists, the storage manager had to treat ID
lists as opaque byte arrays. From the storage manager's perspective, when the content of an ID list
changed, the entire list had changed. For a single ID that was inserted or deleted from an ID list,
the corresponding number of bytes written to the transaction log was the maximum configured size for
that ID list, about 8 kilobytes. Also, every database page on which the list was stored was marked as
dirty, since the entire list had changed.
In the redesigned index, the storage manager has visibility into the fine-grain index structure, which
optimizes transaction logging so that only the number of bytes actually changed need to be logged for
any given index modification. T he Berkeley DB provides ID list semantics, which are implemented by the
storage manager. T he Berkeley API was enhanced to support the insertion and deletion of individual IDs
stored against a common key, with support for duplicate keys, and an optimized mechanism for the
retrieval of the complete ID list for a given key.
T he storage manager has direct knowledge of the application's intent when changes are made to ID
lists, resulting in several improvements to ID list handling:
For long ID lists, the number of bytes written to the transaction log for any update to the list is
significantly reduced, from the maximum ID list size (8 kilobytes) to twice the size of one ID (4 bytes).
For short ID lists, storage efficiency, and in most cases performance, is improved because only the
storage manager meditate need to be stored, not the ID list metadata.
T he average number of database pages marked as dirty per ID insert or delete operation is very
small because a large number of duplicate keys will fit into each database page.
229
Red Hat D irectory Server 8.2 Administration Guide
7.1.5.2. Search Performance
For each entry ID list, there is a size limit that is globally applied to all index keys managed by the server.
In previous versions of Directory Server, this limit was called the All IDs Threshold. Because maintaining
large ID lists in memory can affect performance, the All IDs T hreshold set a limit on how large a single
entry ID list could get. When a list hit a certain pre-determined size, the search would treat it as if the
index contained the entire directory.
T he difficulty in setting the All IDs T hreshold hurt performance. If the threshold was too low, too many
searches examined every entry in the directory. If it was too high, too many large ID lists had to be
maintained in memory.
T he problems addressed by the All IDs T hreshold are no longer present because of the efficiency of
entry insertion, modification, and deletion in the Berkeley DB design. T he All IDs T hreshold is removed
for database write operations, and every ID list is now maintained accurately.
Since loading a long ID list from the database can significantly reduce search performance, the
configuration parameter, nsslapd-idlistscanlim it, sets a limit on the number of IDs that are read
before a key is considered to match the entire primary index. nsslapd-idlistscanlim it is
analogous to the All IDs T hreshold, but it only applies to the behavior of the server's search code, not
the content of the database.
When the server uses indexes in the processing of a search operation, it is possible that one index key
matches a large number of entries. For example, consider a search for
objectclass=inetorgperson in a directory that contained one million inetorgperson entries.
When the server reads the inetorgperson index key in the objectclass index, it finds one million
matching entries. In cases like this, it is more efficient simply to conclude in the index lookup phase of the
search operation processing that all the entries in the database match the query. T his causes
subsequent search processing to scan the entire database content, checking each entry as to whether
it matches the search filter. T he time required to fetch the index keys is not worthwhile; the search
operation is likely to be processed more efficiently by omitting the index lookup.
In Directory Server, when examining an index, if more than a certain number of entries are found, the
server stops reading the index and marks the search as unindexed for that particular index.
T he threshold number of entries is called the idlistscanlim it and is configured with the nsslapdidlistscanlim it configuration attribute. T he default value is 4 000, which is designed to give good
performance for a common range of database sizes and access patterns. T ypically, it is not necessary
to change this value. However, in rare circumstances it may be possible to improve search performance
with a different value. For example, lowering the value will improve performance for searches that will
otherwise eventually hit the default limit of 4000. T his might reduce performance for other searches that
benefit from indexing. Conversely, increasing the limit could improve performance for searches that were
previously hitting the limit. With a higher limit, these searches could benefit from indexing where
previously they did not.
For more information on search limits for the server, refer to Section 7.1.3, “Overview of the Searching
Algorithm”.
7.1.5.3. Backwards Compatibility and Migration
While current versions of Directory Server can support the old database design, only the new design is
supported for this and later releases of Directory Server.
Upon startup, the server will read the database version from the DBVERSION file, which contains the text
Netscape-ldbm /6.2 (old database version), Netscape-ldbm /7.1 (new database format), or
bdb/4 .2/libback-ldbm (new database format). If the file indicates that the old format is used, then
230
Chapter 7. Managing Indexes
the old code is selected for the database. Because the DBVERSION file stores everything per-backend,
it is possible to have different database formats for different individual backends, but the old database
format is not recommended.
All databases must be migrated to Directory Server 8.2 when the system is upgraded. Migration is
supported for Directory Server 6.x versions, and for releases earlier than version 6.x, dump the
databases be dumped, and install Directory Server fresh. Migrating databases is covered in the
Directory Server Installation Guide.
Also, the index sizes can be larger than in older releases, so you may want to increase your database
cache size. T o reconfigure your cache size, look up the nsslapd-dbcachesize entry in the Directory
Server Configuration and Command-Line Tool Reference.
7.1.6. Balancing the Benefits of Indexing
Before creating new indexes, balance the benefits of maintaining indexes against the costs.
Approximate indexes are not efficient for attributes commonly containing numbers, such as telephone
numbers.
Substring indexes do not work for binary attributes.
Equality indexes should be avoided if the value is big (such as attributes intended to contain
photographs or passwords containing encrypted data).
Maintaining indexes for attributes not commonly used in a search increases overhead without
improving global searching performance.
Attributes that are not indexed can still be specified in search requests, although the search
performance may be degraded significantly, depending on the type of search.
T he more indexes you maintain, the more disk space you require.
Indexes can become very time-consuming. For example:
1. T he Directory Server receives an add or modify operation.
2. T he Directory Server examines the indexing attributes to determine whether an index is
maintained for the attribute values.
3. If the created attribute values are indexed, then the Directory Server generates the new index
entries.
4. Once the server completes the indexing, the actual attribute values are created according to the
client request.
For example, the Directory Server adds the entry:
dn: cn=John Doe,ou=People,dc=example,dc=com
objectclass: top
objectClass: person
objectClass: orgperson
objectClass: inetorgperson
cn: John Doe
cn: John
sn: Doe
ou: Manufacturing
ou: people
telephoneNumber: 408 555 8834
description: Manufacturing lead for the Z238 line of widgets.
T he Directory Server is maintaining the following indexes:
231
Red Hat D irectory Server 8.2 Administration Guide
Equality, approximate, and substring indexes for cn (common name) and sn (surname) attributes.
Equality and substring indexes for the telephone number attribute.
Substring indexes for the description attribute.
When adding that entry to the directory, the Directory Server must perform these steps:
1. Create the cn equality index entry for John and John Doe.
2. Create the appropriate cn approximate index entries for John and John Doe.
3. Create the appropriate cn substring index entries for John and John Doe.
4. Create the sn equality index entry for Doe.
5. Create the appropriate sn approximate index entry for Doe.
6. Create the appropriate sn substring index entries for Doe.
7. Create the telephone number equality index entry for 4 08 555 8834 .
8. Create the appropriate telephone number substring index entries for 4 08 555 8834 .
9. Create the appropriate description substring index entries for Manufacturing lead for the
Z238 line of widgets. A large number of substring entries are generated for this string.
As this example shows, the number of actions required to create and maintain databases for a large
directory can be resource-intensive.
7.2. Creating Standard Indexes
T his section describes how to create presence, equality, approximate, substring, and international
indexes for specific attributes using the Directory Server Console and the command line.
When a new index type is created, that index is used as a template for any additional databases as they
are added. T he Directory Server uses the current set of default indexes defined for the instance as the
basis for additional databases.
However, new indexes are not added to existing databases automatically, though they can be added
manually. T his means that if you add a default index to your second database instance, it will not be
maintained in your first database instance but will be maintained in any subsequent instances. T o apply
a new index to an existing database, run the db2index.pl script or run a cn=index,cn=tasks task,
as described in Section 7.3, “Applying New Indexes to Existing Databases”.
Section 7.2.1, “Creating Indexes from the Server Console”
Section 7.2.2, “Creating Indexes from the Command Line”
7.2.1. Creating Indexes from the Server Console
T o create presence, equality, approximate, substring, or international indexes:
1. Select the Configuration tab.
2. Expand the Data node, expand the suffix of the database to index, and select the database.
3. Select the Indexes tab in the right pane.
232
Chapter 7. Managing Indexes
NOTE
Do not click the Database Settings node because this opens the Default Index
Settings window, not the window for configuring indexes per database.
4. If the attribute to be indexed is listed in the Additional Indexes table, go to step 6. Otherwise,
click Add Attribute to open a dialog box with a list of all of the available attributes in the server
schema.
5. Select the attribute to index, and click OK.
T he server adds the attribute to the Additional Indexes table.
6. Select the checkbox for each type of index to maintain for each attribute.
233
Red Hat D irectory Server 8.2 Administration Guide
7. T o create an index for a language other than English, enter the OID of the collation order to use in
the Matching Rules field.
T o index the attribute using multiple languages, list multiple OIDs separated by commas, but no
whitespace. For a list of languages, their associated OIDs, and further information regarding
collation orders, see Appendix C, Internationalization.
8. Click Save.
T he new index is immediately active for any new data that you add and any existing data in your
directory. You do not have to restart your server.
7.2.2. Creating Indexes from the Command Line
NOTE
You cannot create new system indexes because system indexes are hard-coded in Directory
Server.
Use ldapm odify
[6 ]
to add the new index attributes to your directory.
T o create a new index that will become one of the default indexes, add the new index attributes to
the cn=default indexes,cn=config,cn=ldbm database,cn=plugins,cn=config
entry.
T o create a new index for a particular database, add it to the
cn=index,cn=database_name,cn=ldbm database,cn=plugins,cn=config entry, where
cn=database_name corresponds to the name of the database.
NOTE
Avoid creating entries under cn=config in the dse.ldif file. T he cn=config entry in the
simple, flat dse.ldif configuration file is not stored in the same highly scalable database as
regular entries. As a result, if many entries, particularly entries that are likely to be updated
frequently, are stored under cn=config, performance will probably suffer. Although we
recommend you do not store simple user entries under cn=config for performance reasons, it
can be useful to store special user entries such as the Directory Manager entry or replication
manager (supplier bind DN) entry under cn=config since this centralizes configuration
information.
For information on the LDIF update statements required to add entries, see Section 3.3, “Using LDIF
Update Statements to Create or Modify Entries”.
For example, to create presence, equality, and substring indexes for the sn (surname) attribute in the
Exam ple1 database:
1. Open the Directory Server LDAP tool directory.
234
Chapter 7. Managing Indexes
cd /usr/lib64/mozldap
2. Run ldapm odify.
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389
-h server.example.com
T he ldapm odify utility binds to the server and prepares it to add an entry to the configuration
file.
3. Add the LDIF entry for the new indexes:
dn: cn=sn,cn=index,cn=Example1,cn=ldbm database,cn=plugins,cn=config
objectClass:top
objectClass:nsIndex
cn:sn
nsSystemIndex:false
nsIndexType:pres
nsIndexType:eq
nsIndexType:sub
nsMatchingRule:2.16.840.1.113730.3.3.2.3.1
T he cn attribute contains the name of the attribute to index, in this example the sn attribute. T he
entry is a member of the nsIndex object class. T he nsSystem Index attribute is false,
indicating that the index is not essential to Directory Server operations. T he multi-valued
nsIndexT ype attribute specifies the presence (pres), equality (eq) and substring (sub)
indexes. Each keyword has to be entered on a separate line. T he nsMatchingRule attribute in
the example specifies the OID of the Bulgarian collation order; the matching rule can indicate any
possible value match, such as languages or other formats like date or integer.
You can use the keyword none in the nsIndexT ype attribute to specify that no indexes are to be
maintained for the attribute. T his example temporarily disables the sn indexes on the Exam ple1
database by changing the nsIndexT ype to none:
dn: cn=sn,cn=index,cn=Example1,cn=ldbm database,cn=plugins,cn=config
objectClass:top
objectClass:nsIndex
cn:sn
nsSystemIndex:false
nsIndexType:none
For a complete list of matching rules and their OIDs, see Section 8.3.4, “Using Matching Rules”, and for
the index configuration attributes or the ldapm odify command-line utility, see the Directory Server
Configuration and Command-Line Tool Reference.
NOTE
Always use the attribute's primary name (not the attribute's alias) when creating indexes. T he
primary name of the attribute is the first name listed for the attribute in the schema; for example,
uid for the user ID attribute.
7.3. Applying New Indexes to Existing Databases
235
Red Hat D irectory Server 8.2 Administration Guide
New indexes are not added to existing databases automatically. T hey must be added manually, and
Directory Server has two methods for applying new indexes to an existing database: running the
db2index.pl script or running a cn=index,cn=tasks task.
7.3.1. Running the db2index.pl Script
After creating an indexing entry or adding additional index types to an existing indexing entry, run the
db2index.pl script to generate the new set of indexes to be maintained by the Directory Server. After
the script is run, the new set of indexes is active for any new data added to the directory and any
existing data in the directory.
1. Open the Directory Server instance directory.
cd /etc/dirsrv/slapd-instance_name
2. Run the db2index.pl Perl script.
db2index.pl-D "cn=Directory Manager" -w secret -n ExampleServer -t sn
For more information about using this Perl script, see the Directory Server Configuration and
Command-Line Tool Reference.
T able 7.6, “db2index Options” describes the db2index.pl options.
T able 7.3. db2index.pl Options
Option
Description
-D
Specifies the DN of the administrative user.
-w
Specifies the password of the administrative user.
-n
Specifies the name of the database being
indexed.
-t
Specifies the name of the attribute contained by
the index.
7.3.2. Using a cn=tasks Entry to Create an Index
T he cn=tasks,cn=config entry in the Directory Server configuration is a container entry for
temporary entries that the server uses to manage tasks. Several common directory tasks have container
entries under cn=tasks,cn=config. T emporary task entries can be created under
cn=index,cn=tasks,cn=config to initiate an indexing operation.
T his task entry requires a unique name (cn) and a definition for the attribute and type of index, set in
nsIndexAttribute in the format attribute:index_type.
For example:
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=example presence index,cn=index,cn=tasks,cn=config
objectclass: extensibleObject
cn: example presence index
nsIndexAttribute: "cn:pres"
236
Chapter 7. Managing Indexes
T here are three possible index_types:
pres for presence indexes
eq for equality indexes
sub for substring indexes
As soon as the task is completed, the entry is removed from the directory configuration.
T he Directory Server Configuration and Command-Line Tool Reference has more information on running
Directory Server tasks under the cn=tasks entries.
7.4. Creating Browsing (VLV) Indexes
A virtual list view (VLV) index is a way of creating a truncated list for faster searching while enhancing
server performance. T he VLV index itself can be resource-intensive to maintain, but it can be beneficial
in large directories (over 1000 entries).
A browsing index is a type of VLV index that organizes the entries listed into alphabetical order, making it
easier to find entries.
VLV indexes are not applied to attributes, like standard indexes are, but they are dynamically generated
based on attributes set in entries and the location of those entries in the directory tree. VLV indexes,
unlike standard indexes, are special entries in the database rather than configuration settings for the
database.
NOTE
VLV indexes are similar to simple paged results, which can be returned with some external LDAP
clients. Simple paged results are calculated per search, while VLV indexes are a permanent list,
so VLV indexes are overall faster for searches, but do require some overhead for the server to
maintain.
Simple paged results and VLV indexes cannot be used on the same search.
For more information, see Section 8.7, “Using Simple Paged Results”.
7.4.1. Creating Browsing Indexes from the Server Console
1. Select the Directory tab.
2. In the left navigation tree, select the entry, such as People, for which to create the index.
3. From the Object menu, select Create Browsing Index.
237
Red Hat D irectory Server 8.2 Administration Guide
T he Create Browsing Index dialog box appears displaying the status of the index creation.
Click the Status Logs box to view the status of the indexes created.
4. Click Close.
T he new index is immediately active for any new data that is added to the directory. You do not have to
restart your server.
For more information on how to change the VLV search information or the access control rules that are
set by default for VLV searches, see Section 7.4.2.1, “Adding a Browsing Index Entry” and Section 7.4.3,
“Setting Access Control for VLV Information”.
7.4.2. Creating Browsing Indexes from the Command Line
Creating a browsing index or virtual list view (VLV) index from the command line has these steps:
1. Using ldapm odify to add new browsing index entries or edit existing browsing index entries.
See Section 7.4.2.1, “Adding a Browsing Index Entry”.
2. Running the vlvindex script to generate the new set of browsing indexes to be maintained by
the server. See Section 7.4.2.2, “Running the vlvindex Script”. Alternatively, launch an appropriate
task under cn=tasks,cn=config (Section 7.4.2.3, “Using a cn=tasks Entry to Create a
Browsing Index”).
3. Ensuring that access control on VLV index information is set appropriately. See Section 7.4.3,
238
Chapter 7. Managing Indexes
“Setting Access Control for VLV Information”.
7.4 .2.1. Adding a Browsing Index Entry
T he type of browsing index entry to create depends on the type of ldapsearch attribute sorting to
accelerate. It is important to take the following into account:
T he scope of the search (base, one, sub)
T he base of the search (the entry to use as a starting point for the search)
T he attributes to sort
T he filter of the search
For more information on specifying filters for searches, see Chapter 8, Finding Directory Entries.
T he LDBM database to which the entry that forms the base of the search belongs. You can only
create browsing indexes in LDBM databases.
T here is more information on ldapsearch options in the Directory Server Configuration and
Command-Line Tool Reference.
For example, create a browsing index to accelerate an ldapsearch on the entry
ou=People,dc=exam ple,dc=com held in the Exam ple1 database with the following attributes:
T he search base is ou=People,dc=exam ple,dc=com
T he search filter is (|(objectclass=* )(objectclass=ldapsubentry))
T he scope is one
T he sorting order for the returned attributes is cn, givenname, o, ou, and sn
1. Run ldapm odify.
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389
-h server.example.com
T he ldapm odify utility binds to the server and prepares it to add an entry to the configuration
file.
2. Add an entry which specifies the base, scope, and filter of the browsing index:
dn: cn=MCC ou=People dc=example dc=com,cn=userRoot,cn=ldbm
database,cn=plugins,cn=config
objectClass: top
objectClass: vlvSearch
cn: MCC ou=People dc=example dc=com
vlvBase: ou=People,dc=example,dc=com
vlvScope: 1
vlvFilter: (|(objectclass=*)(objectclass=ldapsubentry))
T he cn contains the browsing index identifier, which specifies the entry on which to create the
browsing index; in this example, the ou=People,dc=exam ple,dc=com entry. Red Hat
recommends using the dn of the entry for the browsing index identifier, which is the approach
adopted by the Directory Server Console, to prevent identical browsing indexes from being
created. T he entry is a member of the vlvSearch object class.
T he vlvbase attribute value specifies the entry on which you want to create the browsing
index; in this example, the ou=People,dc=exam ple,dc=com entry (the browsing index
identifier).
T he vlvscope attribute is 1, indicating that the scope for the search you want to accelerate is
239
Red Hat D irectory Server 8.2 Administration Guide
1. A search scope of 1 means that only the immediate children of the entry specified in the cn
attribute, and not the entry itself, will be searched.
T he vlvfilter specifies the filter to be used for the search; in this example,
(|(objectclass=* )(objectclass=ldapsubentry)).
3. Add the second entry, to specify the sorting order for the returned attributes:
dn: cn=by MCC ou=People dc=example dc=com,cn=MCC ou=People
dc=example dc=com,cn=userRoot,cn=ldbm database,cn=plugins,
cn= config
objectClass: top
objectClass: vlvIndex
cn: by MCC ou=People dc=example dc=com
vlvSort: cn givenName o ou sn
T he cn contains the browsing index sort identifier. T he above cn is the type created by the
Console by default, which has the sorting order as being set by the browsing index base. T he
entry is a member of the vlvIndex object class.
T he vlvsort attribute value specifies the order in which you want your attributes to be sorted;
in this example, cn, givenName, o, ou, and then sn.
NOTE
T his first browsing index entry must be added to the cn=database_name,cn=ldbm
database,cn=plugins,cn=config directory tree node, and the second entry must be a
child of the first entry.
7.4 .2.2. Running the vlvindex Script
After creating the two browsing indexing entries or added additional attribute types to an existing
indexing browsing entries, run the vlvindex script to generate the new set of browsing indexes to be
maintained by the Directory Server. After running the script, the new set of browsing indexes is active for
any new data added to the directory and any existing data in the directory.
T o run the vlvindex script:
1. Open the Directory Server instance directory.
cd /etc/dirsrv/slapd-instance_name
2. Stop the server.
service dirsrv stop instance
3. Run the vlvindex script.
vlvindex -n Example1 -T "by MCC ou=people dc=example dc=com"
For more information about using this script, see the Directory Server Configuration and
Command-Line Tool Reference.
4. Restart the server.
24 0
Chapter 7. Managing Indexes
T able 7.4, “vlvindex Options” describes the vlvindex options used in the examples:
T able 7.4 . vlvindex Options
Option
Description
-n
Name of the database containing the entries to
index.
-T
Browsing index identifier to use to create
browsing indexes.
7.4 .2.3. Using a cn=tasks Entry to Create a Browsing Index
As an alternative to running the vlvindex script, it is possible to initiate an indexing task directly.
NOTE
Running the indexing task is the same as running the vlvindex script.
T he cn=tasks,cn=config entry in the Directory Server configuration is a container entry for
temporary entries that the server uses to manage tasks. Several common directory tasks have container
entries under cn=tasks,cn=config. T emporary task entries can be created under
cn=index,cn=tasks,cn=config to initiate an indexing operation.
T his task entry requires a unique name (cn) and one other attribute, nsIndexVLVAttribute, which
gives the name of the browsing index definition entry to use to generate the VLV index.
For example:
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=example VLV index,cn=index,cn=tasks,cn=config
objectclass: extensibleObject
cn: example VLV index
nsIndexVLVAttribute: "by MCC ou=people,dc=example,dc=com"
As soon as the task is completed, the entry is removed from the directory configuration.
T he Directory Server Configuration and Command-Line Tool Reference has more information on running
Directory Server tasks under the cn=tasks entries.
7.4.3. Setting Access Control for VLV Information
T he default access control for the VLV index information is to allow anyone who has authenticated. If a
site requires anonymous users to use the VLV index information, modify the access control set for cn:
VLV Request Control in the Directory Server's configuration.
1. Open the Directory Server configuration directory:
cd /etc/dirsrv/slapd-instance_name
2. In a text editor, open the dse.ldif file.
24 1
Red Hat D irectory Server 8.2 Administration Guide
3. Locate the oid=2.16.84 0.1.113730.3.4 .9 entry.
dn: oid=2.16.840.1.113730.3.4.9,cn=features,cn=config
objectClass: top
objectClass: directoryServerFeature
oid: 2.16.840.1.113730.3.4.9
cn: VLV Request Control
aci: (targetattr != "aci")(version 3.0; acl "VLV Request Control";
allow( read, search, compare, proxy ) userdn = "ldap:///all" ;)
creatorsName: cn=server,cn=plugins,cn=config modifiersName:
cn=server,cn=plugins,cn=config ...
4. Change "ldap://all" to "ldap://anyone" and save your changes.
7.5. Changing the Index Sort Order
By default, indexes are sorted alphabetically, in descending ASCII order. T his is true for every attribute,
even attributes which may have numeric attribute values like Integer or T elephoneNumber. It is possible
to change the sort method by changing the matching rule set for the attribute.
7.5.1. Changing the Sort Order in the Console
1. Select the Configuration tab.
2. Expand the Data node, expand the suffix of the database to index, and select the database.
3. Select the Indexes tab in the right pane.
4. Select the index, and, in the Matching Rules field, enter the new sort order to use. For example,
to sort by numbers, rather than alphabetically, enter integerOrderingMatch.
5. Click Save.
7.5.2. Changing the Sort Order in the Command Line
T o change the sort order using the command line, change the nsMatchingRule for the attribute index.
For example:
24 2
Chapter 7. Managing Indexes
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389
dn: cn=sn,cn=index,cn=Example1,cn=ldbm database,cn=plugins,cn=config
changetype:modify
replace:nsMatchingRule
nsMatchingRule:integerOrderingMatch
7.6. Changing the Width for Indexed Substring Searches
By default, for a search to be indexed, the search string must be at least three characters long, without
counting any wildcard characters. For example, the string abc would be an indexed search while ab*
would not be. Indexed searches are significantly faster than unindexed searches, so changing the
minimum length of the search key is helpful to increase the number of indexed searches.
T o improve search performance, particularly for sites with many wildcard searches, the search string
length for indexed searches can be changed. Directory Server has three attributes which allow you to
change the minimum number of characters required for an indexed search:
T he nsSubStrBegin attribute sets the required number of characters for an indexed search for the
beginning of a search string, before the wildcard.
abc*
T he nsSubStrMiddle attribute sets the required number of characters for an indexed search where
a wildcard is used in the middle of a search string. For example:
ab*z
T he nsSubStrEnd attribute sets the required number of characters for an indexed search for the
end of a search string, after the wildcard. For example:
*xyz
T he default substring search length for the string triplet (before, middle, and end) is 3, 3, and 3, meaning
every search requires a minimum of three characters, in every wildcard position.
For any attribute index to have alternate string lengths, add the extensibleObject object class to the
entry and then set the substring search lengths.
1. Set the new key length for the specific attribute index. T his requires adding the
extensibleObject object class and then adding the nsSubStrBegin, nsSubStrEnd, or
nsSubStrMiddle attributes as appropriate. For example:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: attribute_name,cn=index,cn=database_name,cn=ldbm
database,cn=plugins,cn=config
objectClass: extensibleObject
nsSubStrBegin: 2
nsSubStrMiddle: 2
nsSubStrEnd: 2
2. Stop the server.
24 3
Red Hat D irectory Server 8.2 Administration Guide
service dirsrv stop
3. Recreate the attribute index. If even one of the substring search width options is changed, then
the entire index must be recreated.
db2index -t attribute_name
4. Start the server again.
service dirsrv start
7.7. Deleting Indexes
T his section describes how to delete presence, equality, approximate, substring, international, and
browsing indexes for specific attributes.
NOTE
Because Directory Server 8.2 can operate in either a single or multi-database environment, you
have to delete any unwanted indexes from every database instance. Any default indexes you
delete will not be deleted from previous sets of indexes on existing database instances.
Section 7.7.1, “Deleting Indexes from the Server Console”
Section 7.7.2, “Deleting Indexes from the Command Line”
Section 7.7.3, “Deleting Browsing Indexes from the Server Console”
Section 7.7.4, “Deleting Browsing Indexes from the Command Line”
WARNING
Do not delete system indexes because deleting them can significantly affect Directory Server
performance. System indexes are located in the cn=index,cn=instance,cn=ldbm
database,cn=plugins,cn=config entry and the cn=default
indexes,cn=config,cn=ldbm database,cn=plugins,cn=config entry.
Also, be cautious when deleting default indexes since this can also affect how Directory Server
works.
7.7.1. Deleting Indexes from the Server Console
T he Directory Server Console can delete any custom indexes, indexes used by other server
applications such as a messaging or web server, and default indexes.
NOTE
You cannot delete system indexes.
1. Select the Configuration tab.
2. Expand the Data node and expand the suffix associated with the database containing the index.
24 4
Chapter 7. Managing Indexes
3. Select the database from which to delete the index.
4. Locate the attribute containing the index to delete. Clear the checkbox under the index.
T o delete all indexes maintained for a particular attribute, select the attribute's cell under
Attribute Nam e, and click Delete Attribute.
5. Click Save.
A Delete Index warning dialog box opens, requiring a confirmation to delete the index.
6. Click Yes to delete the index.
7.7.2. Deleting Indexes from the Command Line
Creating browsing indexes, or virtual list view (VLV) indexes, using the ldapdelete command-line
utility has two steps:
1. Delete an entire index entry or delete unwanted index types from an existing index entry using the
ldapdelete command-line utility (Section 7.7.2.1, “Deleting an Index Entry”).
2. Generate the new set of indexes to be maintained by the server using the db2index.pl Perl
script (Section 7.7.2.2, “Running the db2index.pl Script”).
7.7.2.1. Deleting an Index Entry
Use the ldapdelete command-line utility to delete either the entire indexing entry or the unwanted
index types from an existing entry.
T o delete the indexes for a particular database, remove the index entry from the
cn=index,cn=database_name,cn=ldbm database,cn=plugins,cn=config entry, where
24 5
Red Hat D irectory Server 8.2 Administration Guide
cn=database_name corresponds to the name of the database.
T o delete a default index, remove it from the cn=default indexes,cn=config,cn=ldbm
database,cn=plugins,cn=config entry.
1. Run ldapdelete.[6 ]
ldapdelete -D "cn=Directory Manager" -w secret -h ExampleServer
-p 389 "cn=sn,cn=index,cn=Example1,dn=ldbm
database,cn=plugins,dn=config"
For complete information on ldapdelete, see the Directory Server Configuration and CommandLine Tool Reference.
2. For example, delete the presence, equality, and substring indexes for the sn attribute on the
database named Exam ple1:
dn: cn=sn,cn=index,cn=Example1,cn=ldbm database,cn=plugins,cn=config
objectClass:top
objectClass:nsIndex
cn:sn
nsSystemIndex:false
nsIndexType:pres
nsIndexType:eq
nsIndexType:sub
nsMatchingRule:2.16.840.1.113730.3.3.2.3.1
T able 7.5, “ldapdelete Options” describes the ldapdelete options.
T able 7.5. ldapdelete Options
Option
Description
-D
Specifies the distinguished name with which to
authenticate to the server. T he value must be a
DN recognized by the Directory Server, and it
must also have the authority to modify the entries.
-w
Specifies the password associated with the
distinguished name specified in the -D option.
-h
Specifies the name of the host on which the
server is running.
-p
Specifies the port number that the server uses.
After deleting the index entry, the presence, equality, and substring indexes for the sn attribute are no
longer maintained by the Exam ple1 database.
7.7.2.2. Running the db2index.pl Script
After deleting an indexing entry or some of the index types from an indexing entry, run the
db2index.pl script to generate the new set of indexes to be maintained by the Directory Server. Once
you run the script, the new set of indexes is active for any new data you add to your directory and any
existing data in your directory.
T o run the db2index.pl Perl script:
1. Open the Directory Server instance directory.
24 6
Chapter 7. Managing Indexes
cd /etc/dirsrv/slapd-instance_name
2. Run the db2index.pl Perl script. For example:
db2index.pl-D "cn=Directory Manager" -w secret -n Example1
For more information about using the db2index.pl Perl script, refer to Directory Server Configuration
and Command-Line Tool Reference. T able 7.6, “db2index Options” describes the db2index.pl
options used in the examples:
T able 7.6. db2index Options
Option
Description
-D
Specifies the DN of the administrative user.
-w
Specifies the password of the administrative user.
-n
Specifies the name of the database into which
you are importing the data.
7.7.3. Deleting Browsing Indexes from the Server Console
1. Select the Directory tab.
2. Select the entry from which to delete the index in the navigation tree, and select Delete
Browsing Index from the Object menu.
Alternatively, select and right-click the entry of the index to delete in the navigation tree, and then
choose Delete Browsing Index from the pop-up menu.
3. A Delete Browsing Index dialog box appears asking you to confirm the deletion of the index.
Click Yes.
4. T he Delete Browsing Index dialog box appears displaying the status of the index deletion.
24 7
Red Hat D irectory Server 8.2 Administration Guide
7.7.4. Deleting Browsing Indexes from the Command Line
Deleting a browsing index or virtual list view (VLV) index from the command line involves two steps:
1. Using the ldapdelete to delete browsing index entries or edit existing browsing index entries
(Section 7.7.4.1, “Deleting a Browsing Index Entry”).
2. Running the vlvindex script to generate the new set of browsing indexes to be maintained by
the server (Section 7.7.4.2, “Running the vlvindex Script”). Alternatively, launch an appropriate task
under cn=tasks,cn=config (Section 7.4.2.3, “Using a cn=tasks Entry to Create a Browsing
Index”).
T he actual entries for an alphabetical browsing index and virtual list view are the same. T he following
sections describe the steps involved in deleting browsing indexes.
7.7.4 .1. Deleting a Browsing Index Entry
Use the ldapdelete command-line utility to either delete browsing indexing entries or edit existing
browsing index entries. T o delete browsing indexes for a particular database, remove the browsing
index entries from the cn=index,cn=database_name,cn=ldbm
database,cn=plugins,cn=config entry, where cn=database_name corresponds to the name of
the database.
For example, there is a browsing index for accelerating ldapsearch operations on the entry
ou=People,dc=exam ple,dc=com . It held in the Exam ple1 database where the search base is
ou=People,dc=exam ple,dc=com , the search filter is
(|(objectclass=* )(objectclass=ldapsubentry)), the scope is 1, and the sorting order for the
returned attributes is cn, givenname, o, ou, and sn.
1. Run ldapdelete.[6 ]
ldapdelete-D "cn=Directory Manager" -w secret -h ExampleServer
-p 389 "cn=MCC ou=People dc=example dc=com,cn=userRoot,cn=ldbm
database,cn=plugins,cn=config"
"cn=by MCC ou=People dc=example dc=com,cn=MCC ou=People
dc=example dc=com,cn=userRoot,cn=ldbm database,cn=plugins,cn=config"
For full information on ldapdelete options, see the Directory Server Configuration and
Command-Line Tool Reference.
2. T o delete this browsing index, delete the two corresponding browsing index entries:
dn: cn=MCC ou=People dc=example dc=com,cn=userRoot,cn=ldbm
database,cn=plugins,cn=config
objectClass: top
objectClass: vlvSearch
cn: MCC ou=People dc=example dc=com
vlvBase: ou=People,dc=example,dc=com
vlvScope: 1 vlvFilter: (|(objectclass=*)(objectclass=ldapsubentry))
dn: cn=by MCC ou=People dc=example dc=com,cn=MCC ou=People
dc=example dc=com,cn=userRoot,cn=ldbm database,cn=plugins,cn=config
objectClass: top
objectClass: vlvIndex
cn: by MCC ou=People dc=example dc=com
vlvSort: cn givenname o ou sn
24 8
Chapter 7. Managing Indexes
T he following table describes the ldapdelete options used in the example:
Option
Description
-D
Specifies the distinguished name with which to
authenticate to the server. T he value must be a
DN recognized by the Directory Server, and it
must also have the authority to modify the entries.
-w
Specifies the password associated with the
distinguished name specified in the -D option.
-h
Specifies the name of the host on which the
server is running.
-p
Specifies the port number that the server uses.
After deleting the two browsing index entries, the browsing index will no longer be maintained by the
Exam ple1 database.
7.7.4 .2. Running the vlvindex Script
After deleting browsing indexing entries or unwanted attribute types from existing browsing indexing
entries, run the vlvindex script to generate the new set of browsing indexes to be maintained by the
Directory Server. After the script is run, the new set of browsing indexes is active for any new data
added to the directory and any existing data in the directory.
1. Open the Directory Server instance directory.
cd /etc/dirsrv/slapd-instance_name
2. Stop the server.
service dirsrv stop instance
3. Run the vlvindex script.
vlvindex -n Example1 -T "by MCC ou=people dc=example dc=com"
For more information about using the vlvindex script, see the Directory Server Configuration
and Command-Line Tool Reference.
4. Restart the server.
service dirsrv start instance
T able 7.4, “vlvindex Options” describes the vlvindex options.
Alternatively, create a new task entry under cn=index,cn=tasks,cn=config to initiate an indexing
operation. T his task entry requires a unique name (cn) and one other attribute, nsIndexVLVAttribute,
which gives the name of the browsing index definition entry to use to generate the VLV index. T his task
is the same as running vlvindex.
For example:
24 9
Red Hat D irectory Server 8.2 Administration Guide
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=example VLV index,cn=index,cn=tasks,cn=config
objectclass: extensibleObject
cn: example VLV index
nsIndexVLVAttribute: "by MCC ou=people,dc=example,dc=com"
As soon as the task is completed, the entry is removed from the directory configuration.
T he Directory Server Configuration and Command-Line Tool Reference has more information on running
Directory Server tasks under the cn=tasks entries.
[6 ] The LDAP to o ls referenc ed in this g uid e are Mo z illa LDAP, ins talled with Direc to ry Server in the /usr/l i b 6 4 /m o zl d ap
d irec to ry o n Red Hat Enterp ris e Linux 5 (6 4-b it); d irec to ries fo r o ther p latfo rms are lis ted in Sec tio n 1.3, “ LDAP To o l Lo c atio ns ” .
Ho wever, Red Hat Enterp ris e Linux s ys tems als o inc lud e LDAP to o ls fro m O p enLDAP. It is p o s s ib le to us e the O p enLDAP
c o mmand s as s ho wn in the examp les , b ut yo u mus t us e the - x arg ument to d is ab le SASL and allo w s imp le authentic atio n.
250
Chapter 8. Finding D irectory Entries
Chapter 8. Finding Directory Entries
Entries in the directory can be searched for and found using any LDAP client. Most clients provide some
form of search interface so that the directory can be searched easily and entry information can be easily
retrieved.
NOTE
Users cannot search the directory unless the appropriate access control has been set in the
directory. For information on setting access control in the directory, see Chapter 12, Managing
Access Control.
8.1. Finding Entries Using the Directory Server Console
Users can browse the Directory tab of the Directory Server Console to see the contents of the
directory tree and search for specific entries in the directory.
Figure 8.1. Browsing Entries in the Directory T ab
Depending on the DN used to authenticate to the directory, this tab displays the contents of the directory
that the user account has access permissions to view. Browse through the contents of the tree, or rightclick an entry, and select Search from the pop-up menu.
251
Red Hat D irectory Server 8.2 Administration Guide
Figure 8.2. Searching for Entries
WARNING
Do not modify the contents of the o=NetscapeRoot suffix using the Directory tab unless
instructed to do so by Red Hat technical support.
8.2. Using ldapsearch
T he ldapsearch command-line utility can locate and retrieve directory entries. T his utility opens a
connection to the specified server using the specified distinguished name and password and locates
entries based on a specified search filter. T he search scope can include a single entry, an entry's
immediate subentries, or an entire tree or subtree.
Search results are returned in LDIF format.
Red Hat Directory Server uses Mozilla LDAP tools, including ldapsearch. T he MozLDAP tools are
installed with Directory Server and are located in the /usr/lib64 /m ozldap directory for Red Hat
Enterprise Linux 5 (64-bit), and in the /usr/lib/m ozldap directory on Red Hat Enterprise Linux 5 (32bit). When running any LDAP command, make sure to use the MozLDAP utilities, otherwise the command
will return errors.
NOTE
For most Linux systems, OpenLDAP tools are already installed in the /usr/bin/ directory.
T hese OpenLDAP tools will not work for Directory Server operations.
T his section contains information about the following topics:
252
Chapter 8. Finding D irectory Entries
Section 8.2.1, “ldapsearch Command-Line Format”
Section 8.2.2, “Commonly Used ldapsearch Options”
Section 8.2.3, “Using Special Characters”
8.2.1. ldapsearch Command-Line Format
T he ldapsearch command must use the following format:
/usr/lib64/mozldap/ldapsearch [optional_options] [optional_search_filter]
[optional_list_of_attributes]
optional_options is a series of command-line options. T hese must be specified before the search
filter, if any are used.
optional_search_filter is an LDAP search filter as described in Section 8.3, “LDAP Search Filters”. Do
not specify a separate search filter if search filters are specified in a file using the -f option.
optional_list_of_attributes is a list of attributes separated by a space. Specifying a list of attributes
reduces the number of attributes returned in the search results. T his list of attributes must appear
after the search filter. For an example, see Section 8.4.6, “Displaying Subsets of Attributes”. If a list of
attributes is not specified, the search returns values for all attributes permitted by the access control
set in the directory (with the exception of operational attributes).
NOTE
For operational attributes to be returned as a result of a search operation, they must be
explicitly specified in the search command. T o retrieve regular attributes in addition to explicitly
specified operational attributes, use an asterisk (*) in the list of attributes in the ldapsearch
command. T o retrieve no attributes, just a list of the matching DNs, use the special attribute
1.1. T his is useful, for example, to get a list of DNs to pass to the ldapdelete command.
8.2.2. Commonly Used ldapsearch Options
T he following table lists the most commonly used ldapsearch command-line options. If a specified
value contains a space ( ), the value should be surrounded by single or double quotation marks, such as
-b "ou=groups,dc=exam ple,dc=com ".
Option
Description
-b
Specifies the starting point for the search. T he
value specified here must be a distinguished
name that currently exists in the database. T his
is optional if the LDAP_BASEDN environment
variable has been set to a base DN. T he value
specified in this option should be provided in
single or double quotation marks. For example:
-b "cn=Barbara Jensen,ou=Product
Development,dc=example,dc=com"
T o search the root DSE entry, specify an empty
string here, such as -b "" .
-D
Specifies the distinguished name with which to
authenticate to the server. T his is optional if
253
Red Hat D irectory Server 8.2 Administration Guide
anonymous access is supported by the server. If
specified, this value must be a DN recognized by
the Directory Server, and it must also have the
authority to search for the entries. For example, D "uid=bjensen,dc=exam ple,dc=com ".
-h
Specifies the hostname or IP address of the
machine on which the Directory Server is
installed. For example, -h
server.exam ple.com . If a host is not
specified, ldapsearch uses the localhost.
NOTE
Directory Server supports both IPv4 and
IPv6 IP addresses.
-l
Specifies the maximum number of seconds to wait
for a search request to complete. For example, -l
300. T he default value for the nsslapdtimelimit attribute is 3600 seconds.
Regardless of the value specified, ldapsearch
will never wait longer than is allowed by the
server's nsslapd-timelimit attribute.
-p
Specifies the T CP port number that the Directory
Server uses. For example, -p 104 9. T he default
is 389. If -Z is used, the default is 636.
-s
Specifies the scope of the search. T he scope can
be one of the following:
base searches only the entry specified in the -b
option or defined by the LDAP_BASEDN
environment variable.
one searches only the immediate children of the
entry specified in the -b option. Only the children
are searched; the actual entry specified in the -b
option is not searched.
sub searches the entry specified in the -b option
and all of its descendants; that is, perform a
subtree search starting at the point identified in
the -b option. T his is the default.
-w
Gives the password associated with the
distinguished name that is specified in the -D
option. If this option is not specified, anonymous
access is used. For example, -w diner892.
-x
Specifies that the search results are sorted on
the server rather than on the client. T his is useful
for sorting according to a matching rule, as with
an international search. In general, it is faster to
sort on the server rather than on the client.
254
Chapter 8. Finding D irectory Entries
-z
Sets the maximum number of entries to return in
response to a search request. For example, -z
1000. Normally, regardless of the value specified
here, ldapsearch never returns more entries
than the number allowed by the server's
nsslapd-sizelimit attribute. However, this
limitation can be overridden by binding as the root
DN when using this command-line argument.
When binding as the root DN, this option defaults
to zero (0). T he default value for the nsslapdsizelimit attribute is 2000 entries.
For detailed information on all ldapsearch utility options, refer to the Directory Server Configuration
and Command-Line Tool Reference.
8.2.3. Using Special Characters
When using the ldapsearch command-line utility, it may be necessary to specify values that contain
characters that have special meaning to the command-line interpreter, such as space ( ), asterisk (*), or
backslash (\). Enclose the value which has the special character in quotation marks (""). For example:
-D "cn=Barbara Jensen,ou=Product Development,dc=example,dc=com"
Depending on the command-line interpreter, use either single or double quotation marks. In general, use
single quotation marks (') to enclose values. Use double quotation marks (") to allow variable
interpolation if there are shell variables. Refer to the operating system documentation for more
information.
8.3. LDAP Search Filters
Search filters select the entries to be returned for a search operation. T hey are most commonly used
with the ldapsearch command-line utility. When using ldapsearch, there can be multiple search
filters in a file, with each filter on a separate line in the file, or a search filter can be specified directly on
the command line.
T he basic syntax of a search filter is:
attribute operator value
For example:
buildingname>=alpha
In this example, buildingname is the attribute, >= is the operator, and alpha is the value. Filters can
also be defined that use different attributes combined together with Boolean operators.
255
Red Hat D irectory Server 8.2 Administration Guide
TIP
When performing a substring search using a matching rule filter, use the asterisk (*) character as
a wildcard to represent zero or more characters.
For example, to search for an attribute value that starts with the letter l and ends with the letter n,
enter a l* n in the value portion of the search filter. Similarly, to search for all attribute values
beginning with the letter u, enter a value of u* in the value portion of the search filter.
T o search for a value that contains the asterisk (*) character, the asterisk must be escaped with
the designated escape sequence, \5c2a. For example, to search for all employees with
businessCategory attribute values of Exam ple* Net product line, enter the following
value in the search filter:
Example\5c2a*Net product line
8.3.1. Using Attributes in Search Filters
T he most basic sort of search looks for the presence of attributes or specific values in entries. T here
are many variations on how to look for attributes in entries. It is possible to check that the attribute
merely exists, to match an exact value, or to list matches against a partial value.
A presence search uses a wild card (an asterisk) to return every entry which has that attribute set,
regardless of value. For example, this returns every entry which has a manager attribute:
"(manager=*)"
It is also possible to search for an attribute with a specific value; this is called an equality search. For
example:
"(cn=babs jensen)"
T his search filter returns all entries that contain the common name Babs Jensen. Most of the time,
equality searches are not case sensitive.
When an attribute has values associated with a language tag, all of the values are returned. T hus, the
following two attribute values both match the "(cn=babs jensen)" filter:
cn: babs jensen
cn;lang-fr: babs jensen
It is also possible to search for a partial match on an attribute value, a substring index. For example:
"(description=*X.500*)"
"(sn=*nderson)"
"(givenname=car*)"
T he length of the substring searches is configured in the substring index itself, as described in
Section 7.6, “Changing the Width for Indexed Substring Searches”.
8.3.2. Using Operators in Search Filters
Operators in search filters set the relationship between the attribute and the given search value. For
256
Chapter 8. Finding D irectory Entries
people searches, operators can be used to set a range, to return a last names within a subset of letters
in the alphabet or employee numbers that come after a certain number.
"(employeeNumber>=500)"
"(sn~=suret)"
"(salary<=150000)"
Operators also enable phonetic and approximate searches, which allow more effective searches with
imperfect information and are particularly useful in internationalized directories.
T he operators that can be used in search filters are listed in T able 8.1, “Search Filter Operators”. In
addition to these search filters, special filters can be specified to work with a preferred language
collation order. For information on how to search a directory with international charactersets, see
Section C.4, “Searching an Internationalized Directory”.
257
Red Hat D irectory Server 8.2 Administration Guide
T able 8.1. Search Filter Operators
Search T ype
Operator
Description
Equality
=
Returns entries containing
attribute values that exactly
match the specified value. For
example, cn=Bob Johnson
Substring
=string* string
Returns entries containing
attributes containing the
specified substring. For
example, cn=Bob*
cn=* Johnson cn=* John*
cn=B* John. T he asterisk (*)
indicates zero (0) or more
characters.
Greater than or equal to
>=
Returns entries containing
attributes that are greater than
or equal to the specified value.
For example, buildingnam e
>= alpha.
Less than or equal to
<=
Returns entries containing
attributes that are less than or
equal to the specified value. For
example, buildingnam e <=
alpha.
Presence
=*
Returns entries containing one
or more values for the specified
attribute. For example, cn=*
telephoneNum ber=*
m anager=* .
Approximate
~=
Returns entries containing the
specified attribute with a value
that is approximately equal to
the value specified in the search
filter. For example, cn~=suret
l~=san fransico could
return cn=sarette l=san
francisco.
8.3.3. Using Compound Search Filters
Multiple search filter components can be combined using Boolean operators expressed in prefix notation
as follows:
(Boolean-operator(filter)(filter)(filter)...)
Boolean-operator is any one of the Boolean operators listed in T able 8.2, “Search Filter Boolean
Operators”.
FOr example, this filter returns all entries that do not contain the specified value:
258
Chapter 8. Finding D irectory Entries
(!(cn=Ray Kultgen))
(!(objectClass=person))
Obviously, compound search filters are most useful when they are nested together into completed
expressions:
(Boolean-operator(filter)((Boolean-operator(filter)(filter)))
T hese compound filters can be combined with other types of searches (approximate, substring, other
operators) to get very detailed results. For example, this filter returns all entries whose organizational
unit is Marketing and whose description field does not contain the substring X.500:
(&(ou=Marketing)(!(description=*X.500*)))
T hat filter can be expanded to return entries whose organizational unit is Marketing, that do not have
the substring X.500, and that have Julie Fulmer or Cindy Z waska as a manager:
(&(ou=Marketing)(!(description=*X.500*))(|(manager=cn=Julie
Fulmer,ou=Marketing,dc=example,dc=com)(manager=cn=Cindy
Zwaska,ou=Marketing,dc=example,dc=com)))
T his filter returns all entries that do not represent a person and whose common name is similar to
printer3b:
(&(!(objectClass=person))(cn~=printer3b))
T able 8.2. Search Filter Boolean Operators
Operator
Symbol
Description
AND
&
All specified filters must be true
for the statement to be true. For
example, (&(filter)(filter)(filter)...).
OR
|
At least one specified filter must
be true for the statement to be
true. For example,
(|(filter)(filter)(filter)...)
NOT
!
T he specified statement must
not be true for the statement to
be true. Only one filter is
affected by the NOT operator.
For example, (!(filter)).
Boolean expressions are evaluated in the following order:
Innermost to outermost parenthetical expressions first.
All expressions from left to right.
8.3.4. Using Matching Rules
A matching rule tells the Directory Server how to compare two values (the value stored in the attribute
and the value in the search filter). A matching rule also defines how to generate index keys. Matching
rules are somewhat related to attribute syntaxes. Syntaxes define the format of an attribute value;
259
Red Hat D irectory Server 8.2 Administration Guide
matching rules define how that format is compared and indexed.
T here are three different types of matching rules:
EQUALIT Y specifies how to compare two values for an equal match. For example, how to handle
strings like “Fred” and “FRED”. Search filters that test for equality (e.g. attribute=value) use the
EQUALIT Y rule. Equality (eq) indexes use the EQUALIT Y rule to generate the index keys. Update
operations use the EQUALIT Y rule to compare values to be updated with values already in an entry.
ORDERING specifies how to compare two values to see if one value is greater or less than another
value. Search filters that set a range (e.g. attribute<=value or attribute>=value) use the ORDERING
rule. An index for an attribute with an ORDERING rule orders the equality values.
SUBST R specifies how to do substring matching. Substring search filters (e.g.
attribute=*partial_string* or attribute=*end_string) use the SUBST R rule. Substring (sub) indexes use
the SUBST R rule to generate the index keys.
IMPORTANT
A matching rule is required in order to support searching or indexing for the corresponding
search filter or index type. For example, an attribute must have an EQUALIT Y matching rule in
order to support equality search filters and eq indexes for that attribute. An attribute must have
both an ORDERING matching rule and an EQUALIT Y matching rule in order to support range
search filters and indexed range searches.
A search operation will be rejected with PROT OCOL_ERROR or UNWILLING_T O_PERFORM if an
attempt is made to use a search filter for an attribute that has no corresponding matching rule.
An extensible matching rule search filter can be used to search for an attribute value with a different
matching rule than the one defined for the attribute. T he matching rule must be compatible with the
syntax of the attribute being searched. For example, to run a case insensitive search for an attribute that
has a case-sensitive matching rule defined for it, specify a case insensitive matching rule in the search
filter.
NOTE
Matching rules are used for searches in internationalized directories, to specify the language
types to use for the results. T his is covered in Section C.4, “Searching an Internationalized
Directory”.
TIP
An index for an attributes uses whatever matching rules are defined for that attribute in its
schema definition. Additional matching rules to use for an index can be configured using the
nsMatchingRule attribute, as in Section 7.2.2, “Creating Indexes from the Command Line”.
T he syntax of the matching rule filter inserts a matching rule name or OID into the search filter:
attr:matchingRule:=value
attr is an attribute belonging to entries being searched, such as cn or mail.
260
Chapter 8. Finding D irectory Entries
matchingRule is a string that contains the name or OID of the rule to use to match attribute values
according to the required syntax.
value is either the attribute value to search for or a relational operator plus the attribute value to
search for. T he syntax of the value of the filter depends on the matching rule format used.
A matching rule is actually a schema element, and, as with other schema elements is uniquely identified
by an object identifier (OID).
Many of the matching rules defined for Red Hat Directory Server relate to language codes and set
internationalized collation orders supported by the Directory Server. For example, the OID
2.16.84 0.1.113730.3.3.2.17.1 identifies the Finnish collation order.
NOTE
Unlike other schema elements, additional matching rules cannot be added to the Directory Server
configuration.
Most of the matching rules list in T able 8.3, “General Syntax Matching Rules” are used for equality
indexes. Matching rules with ordering in their name are used for ordering indexes, and those with
substring in their name are used for substring (SUBST R) indexes. (T he matching rules used for
international matching and collation orders use a different naming scheme.)
261
Red Hat D irectory Server 8.2 Administration Guide
T able 8.3. General Syntax Matching Rules
Matching Rule
Object Identifiers
(OIDs)
Definitions
Compatible
Syntaxes
Bitwise AND Match
1.2.840.113556.1.4.803
Performs bitwise AND
matches.
T ypically used with: [a]
Integer
Numeric String
Bitwise OR Match
1.2.840.113556.1.4.804
Performs bitwise OR
matches.
T ypically used with:[a]
Integer
Numeric String
generalizedT imeMatch
2.5.13.27
Compares values that
are in a Generalized
T ime format.
Generalized T ime
generalizedT imeOrderi
ngMatch
2.5.13.28
Allows ranged
searches (less than
and greater than) on
values that are in a
Generalized T ime
format.
Generalized T ime
integerMatch
2.5.13.14
Evaluates integer
values.
Integer
integerOrderingMatch
2.5.13.15
Allows ranged
searches (less than
and greater than) on
integer values.
Integer
numericStringMatch
2.5.13.8
Compares more
general numeric values.
Numeric String
numericStringOrdering
Match
2.5.13.9
Allows ranged
searches (less than
and greater than) on
more general numeric
values.
Numeric String
numericStringSubstring
Match
2.5.13.10
Compares more
general numeric values.
Numeric String
[a] This has a s p ec ial fo rmat; the value is c o nverted to integ er b efo re b eing us ed b y Direc to ry Server.
262
Chapter 8. Finding D irectory Entries
T able 8.4 . Language Ordering Matching Rules
Matching Rule
Object Identifiers (OIDs)
English (Case Exact Ordering Match)
2.16.840.1.113730.3.3.2.11.3
Albanian (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.44.1
Arabic (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.1.1
Belorussian (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.2.1
Bulgarian (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.3.1
Catalan (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.4.1
Chinese - Simplified (Case Insensitive Ordering
Match)
2.16.840.1.113730.3.3.2.49.1
Chinese - T raditional (Case Insensitive Ordering
Match)
2.16.840.1.113730.3.3.2.50.1
Croatian (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.22.1
Czechoslovakian (Case Insensitive Ordering
Match)
2.16.840.1.113730.3.3.2.5.1
Danish (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.6.1
Dutch (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.33.1
Dutch - Belgian (Case Insensitive Ordering
Match)
2.16.840.1.113730.3.3.2.34.1
English - US (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.11.1
English - Canadian (Case Insensitive Ordering
Match)
2.16.840.1.113730.3.3.2.12.1
English - Irish (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.14.1
Estonian (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.16.1
Finnish (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.17.1
French (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.18.1
French - Belgian (Case Insensitive Ordering
Match)
2.16.840.1.113730.3.3.2.19.1
French - Canadian (Case Insensitive Ordering
Match)
2.16.840.1.113730.3.3.2.20.1
French - Swiss (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.21.1
German (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.7.1
German - Austrian (Case Insensitive Ordering
Match)
2.16.840.1.113730.3.3.2.8.1
German - Swiss (Case Insensitive Ordering
Match)
2.16.840.1.113730.3.3.2.9.1
Greek (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.10.1
Hebrew (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.27.1
Hungarian (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.23.1
Icelandic (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.24.1
Italian (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.25.1
Italian - Swiss (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.26.1
Japanese (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.28.1
Korean (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.29.1
Latvian, Lettish (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.31.1
263
Red Hat D irectory Server 8.2 Administration Guide
Lithuanian (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.30.1
Macedonian (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.32.1
Norwegian (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.35.1
Norwegian - Bokmul (Case Insensitive Ordering
Match)
2.16.840.1.113730.3.3.2.36.1
Norwegian - Nynorsk (Case Insensitive Ordering
Match)
2.16.840.1.113730.3.3.2.37.1
Polish (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.38.1
Romanian (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.39.1
Russian (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.40.1
Serbian - Cyrillic (Case Insensitive Ordering
Match)
2.16.840.1.113730.3.3.2.45.1
Serbian - Latin (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.41.1
Slovakian (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.42.1
Slovenian (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.43.1
Spanish (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.15.1
Swedish (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.46.1
T urkish (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.47.1
Ukrainian (Case Insensitive Ordering Match)
2.16.840.1.113730.3.3.2.48.1
264
Chapter 8. Finding D irectory Entries
T able 8.5. Language Substring Matching Rules
Matching Rule
Object Identifiers (OIDs)
English (Case Exact Substring Match)
2.16.840.1.113730.3.3.2.11.3.6
Albanian (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.44.1.6
Arabic (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.1.1.6
Belorussian (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.2.1.6
Bulgarian (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.3.1.6
Catalan (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.4.1.6
Chinese - Simplified (Case Insensitive Substring
Match)
2.16.840.1.113730.3.3.2.49.1.6
Chinese - T raditional (Case Insensitive Substring
Match)
2.16.840.1.113730.3.3.2.50.1.6
Croatian (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.22.1.6
Czechoslovakian (Case Insensitive Substring
Match)
2.16.840.1.113730.3.3.2.5.1.6
Danish (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.6.1.6
Dutch (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.33.1.6
Dutch - Belgian (Case Insensitive Substring
Match)
2.16.840.1.113730.3.3.2.34.1.6
English - US (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.11.1.6
English - Canadian (Case Insensitive Substring
Match)
2.16.840.1.113730.3.3.2.12.1.6
English - Irish (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.14.1.6
Estonian (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.16.1.6
Finnish (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.17.1.6
French (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.18.1.6
French - Belgian (Case Insensitive Substring
Match)
2.16.840.1.113730.3.3.2.19.1.6
French - Canadian (Case Insensitive Substring
Match)
2.16.840.1.113730.3.3.2.20.1.6
French - Swiss (Case Insensitive Substring
Match)
2.16.840.1.113730.3.3.2.21.1.6
German (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.7.1.6
German - Austrian (Case Insensitive Substring
Match)
2.16.840.1.113730.3.3.2.8.1.6
German - Swiss (Case Insensitive Substring
Match)
2.16.840.1.113730.3.3.2.9.1.6
Greek (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.10.1.6
Hebrew (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.27.1.6
Hungarian (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.23.1.6
Icelandic (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.24.1.6
Italian (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.25.1.6
Italian - Swiss (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.26.1.6
Japanese (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.28.1.6
Korean (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.29.1.6
265
Red Hat D irectory Server 8.2 Administration Guide
Latvian, Lettish (Case Insensitive Substring
Match)
2.16.840.1.113730.3.3.2.31.1.6
Lithuanian (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.30.1.6
Macedonian (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.32.1.6
Norwegian (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.35.1.6
Norwegian - Bokmul (Case Insensitive Substring
Match)
2.16.840.1.113730.3.3.2.36.1.6
Norwegian - Nynorsk (Case Insensitive Substring
Match)
2.16.840.1.113730.3.3.2.37.1.6
Polish (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.38.1.6
Romanian (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.39.1.6
Russian (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.40.1.6
Serbian - Cyrillic (Case Insensitive Substring
Match)
2.16.840.1.113730.3.3.2.45.1.6
Serbian - Latin (Case Insensitive Substring
Match)
2.16.840.1.113730.3.3.2.41.1.6
Slovakian (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.42.1.6
Slovenian (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.43.1.6
Spanish (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.15.1.6
Swedish (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.46.1.6
T urkish (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.47.1.6
Ukrainian (Case Insensitive Substring Match)
2.16.840.1.113730.3.3.2.48.1.6
8.4. Examples of Common ldapsearches
T he next set of examples assumes the following:
T he search is for all entries in the directory.
T he directory is configured to support anonymous access for search and read. T his means that no
bind information has to be supplied in order to perform the search. For more information on
anonymous access, see Section 12.4.2, “Defining User Access - userdn Keyword”.
T he server is located on a host named server.exam ple.com .
T he server uses port number 389. Since this is the default port, the port number does not have to be
sent in the search request.
SSL is enabled for the server on port 636 (the default SSL port number).
T he suffix under which all data are stored is dc=exam ple,dc=com .
8.4.1. Returning All Entries
Given the previous information, the following call will return all entries in the directory (subject to the
configured size and time resource limits):
/usr/lib64/mozldap/ldapsearch -D "cn=directory manager" -w secret -p 389 -h
server.example.com -b "dc=example,dc=com" -s sub "(objectclass=*)"
"objectclass=* " is a search filter that matches any entry in the directory. Since every entry must
have an object class, and the objectclass attribute is always indexed, this is a useful search filter to
return every entry.
266
Chapter 8. Finding D irectory Entries
8.4.2. Specifying Search Filters on the Command Line
A search filter can be specified directly on the command line as long as the filter is enclosed in quotation
marks ("filter"). If the filter is supplied with the command, do not specify the -f option. For example:
/usr/lib64/mozldap/ldapsearch -D "cn=directory manager" -w secret -p 389 -h
server.example.com -b "dc=example,dc=com" -s sub "cn=babs jensen"
8.4.3. Searching the Root DSE Entry
T he root DSE is a special entry that contains a list of all the suffixes supported by the local Directory
Server. T his entry can be searched by supplying a search base of "", a search scope of base, and a
filter of "objectclass=* ". For example:
/usr/lib64/mozldap/ldapsearch -D "cn=directory manager" -w secret -p 389 -h
server.example.com -b "" -s base "objectclass=*"
8.4.4. Searching the Schema Entry
Directory Server stores all directory server schema in the special cn=schem a entry. T his entry contains
information on every object class and attribute defined for the Directory Server. T he following command
searches the contents of the cn=schem a entry:
/usr/lib64/mozldap/ldapsearch -D "cn=directory manager" -w secret -p 389 -h
server.example.com -b "cn=schema" -s base "objectclass=*"
8.4.5. Using LDAP_BASEDN
T o make searching easier, it is possible to set the search base using the LDAP_BASEDN environment
variable. Doing this means that the search base does not have to be set with the -b option. For
information on how to set environment variables, see the documentation for the operating system.
T ypically, set LDAP_BASEDN to the directory's suffix value. Since the directory suffix is equal to the root,
or topmost, entry in the directory, this causes all searches to begin from the directory's root entry.
For example, set LDAP_BASEDN to dc=exam ple,dc=com and search for cn=babs jensen in the
directory, use the following command-line call:
export LDAP_BASEDN="dc=example,dc=com"
/usr/lib64/mozldap/ldapsearch -D "cn=directory manager" -w secret -p 389 -h
server.example.com "cn=babs jensen"
In this example, the default scope of sub is used because the -s option was not used to specify the
scope.
8.4.6. Displaying Subsets of Attributes
T he ldapsearch command returns all search results in LDIF format. By default, ldapsearch returns
the entry's distinguished name and all of the attributes that a user is allowed to read. T he directory
access control can be set such that users are allowed to read only a subset of the attributes on any
given directory entry. Only operational attributes are not returned. For operational attributes to be
returned as a result of a search operation, explicitly specify them in the search command.
267
Red Hat D irectory Server 8.2 Administration Guide
It may not be necessary to have all of the attributes for an entry returned in the search results. T he
returned attributes can be limited to just a few specific attributes by specifying the desired ones on the
command line immediately after the search filter. For example, to show the cn and sn attributes for every
entry in the directory, use the following command-line call:
/usr/lib64/mozldap/ldapsearch -D "cn=directory manager" -w secret -p 389 -h
server.example.com -b "dc=example,dc=com" -s sub "(objectclass=*)" sn cn
8.4.7. Searching for Operational Attributes
Operational attributes are special attributes set by the Directory Server itself that are used by the server
to perform maintenance tasks, like processing access control instructions. T hey also show specific
information about the entry, like the time it was initially created and the name of the user who created it.
Operational attributes are available for use on every entry in the directory, regardless of whether the
attribute is specifically defined for the object class of the entry.
Operational attributes are not returned in regular ldapsearches, so to return operational attributes,
they have to be explicitly specified in the ldapsearch request.
/usr/lib64/mozldap/ldapsearch -D "cn=directory manager" -w secret -p 389 -h
server.example.com -b "dc=example,dc=com" -s sub "(objectclass=*)" creatorsName
createTimestamp modifiersName modifyTimestamp
T he complete list of operational attributes is in the "Operational Attributes and Object Classes" chapter
in the Schema Reference
8.4.8. Specifying Search Filters Using a File
Search filters can be entered into a file instead of entering them on the command line. In this case,
specify each search filter on a separate line in the file. T he ldapsearch command runs each search in
the order in which it appears in the file.
For example:
sn=Francis
givenname=Richard
ldapsearch first finds all the entries with the surname Francis, then all the entries with the
givenname Richard. If an entry is found that matches both search criteria, then the entry is returned
twice.
For example, in this search, the filters are specified in a file named searchdb:
/usr/lib64/mozldap/ldapsearch -D "cn=directory manager" -w secret -p 389 -h
server.example.com -f searchdb
T he set of attributes returned here can be limited by specifying the attribute names at the end of the
search line. For example, the following ldapsearch command performs both searches but returns only
the DN and the givenname and sn attributes of each entry:
/usr/lib64/mozldap/ldapsearch -D "cn=directory manager" -w secret -p 389 -h
server.example.com -f searchdb sn givenname
8.4.9. Specifying DNs That Contain Commas in Search Filters
268
Chapter 8. Finding D irectory Entries
When a DN within a search filter contains a comma as part of its value, the comma must be escaped with
a backslash (\). For example, to find everyone in the exam ple.com Bolivia, S.A. subtree, use the
following command:
/usr/lib64/mozldap/ldapsearch -D "cn=directory manager" -w secret -p 389 -h
server.example.com -s base -b "l=Bolivia\,S.A.,dc=example,dc=com" "objectclass=*"
8.4.10. Using Client Authentication When Searching
Client authentication uses a stored certificate to bind to the directory rather than simple usernamepassword. T his example shows user bjensen searching the directory using client authentication:
/usr/lib64/mozldap/ldapsearch -h server.example.com -p 636 -b
"dc=example,dc=com" -N "bjensenscertname" -Z -W certdbpassword -P
/home/bjensen/certdb/cert8.db "givenname=Richard"
8.4.11. Searching with Specified Controls
T he Directory Server has defined controls in its supportedControls attribute in its DSE. Some of these
define server operations like replication; other are allowed extended operations like get effective rights
or dereferencing controls which clients can pass through LDAP operations to the server. T hese controls
can be specified using the -J option by giving the control OID, its criticality for the ldapsearch, and
any information required for the control operation.
-J control_OID:boolean criticality:control_information
T his option is used, for example, for Get effective rights searches:
/usr/lib64/mozldap/ldapsearch -D "cn=directory manager" -w secret -p 389 -h
server.example.com -b "dc=example,dc=com" -s sub -J
1.3.6.1.4.1.42.2.27.9.5.2:true:dn:uid=jsmith,ou=people,dc=example,dc=com
"(objectclass=*)"
Get effective rights searches are covered in much more detail in the access control chapter,
Section 12.7, “Checking Access Rights on Entries (Get Effective Rights)”.
8.4.12. Searching with Language Matching Rules
T o explicitly submit a matching rule in a search filter, insert the matching rule after the attribute:
attr:matchingRule:=value
Matching rules are frequently used for searching internationalized directories. For example, this
seearches for the department numbers after N4709 in the Swedish
(2.16.84 0.1.113730.3.3.2.4 6.1) matching rule.
departmentNumber:2.16.840.1.113730.3.3.2.46.1:=>= N4709
More examples of performing internaionalized searches are given in Section C.4, “Searching an
Internationalized Directory”.
8.4.13. Searching for Attributes with Bit Field Values
Bitwise searches use the bitwise AND or bitwise OR matching rules to perform bitwise search operaions
269
Red Hat D irectory Server 8.2 Administration Guide
on attributes with values that are bit fields.
NOTE
Attributes with values for bit fields aren't common in LDAP. (No default Directory Server schema
use bit fields as attribute syntax.) However, several LDAP syntaxes support interger-style values.
Custom attributes can be defined which use bit field values, and applications can use those
custom attributes to perform bitwise operations against bit field values.
T he bitwise AND matching rule (1.2.84 0.113556.1.4 .803) checks that the bit given in the
assertion value is set in the bit field attribute value. (T his is somewhat analogous to an equality search.)
In this example, the userAccountControl value must be set to the bit representing 2.
"(UserAccountControl:1.2.840.113556.1.4.803:=2)"
In this example, the userAccountControl value must have all of the bits set that are set in the value 6
(bits 2 and 4).
"(UserAccountControl:1.2.840.113556.1.4.803:=6)”
T he bitwise OR matching rule (1.2.84 0.113556.1.4 .804 ) checks to see if any of the bits in the
assertion string are represented in the attribute value. (T his is somewhat analogous to a substring
search.) In this example, the userAccountControl value must have any of the bits which are set in the bit
field of 6, meaning that the attribute value can be 2, 4, or 6.
"(UserAccountControl:1.2.840.113556.1.4.804:=6)"
Bitwise searches can be used with Windows-Red Hat Enterprise Linux integration, such as using Samba
file servers.
NOTE
Microsoft has good documentation on bitwise operators at http://msdn.microsoft.com/enus/library/aa746475.
8.5. Using Persistent Search
A persistent search is an ldapsearch which remains open even after the initial search results are
returned.
Persistent searches are especially useful for applications or clients which access the Directory Server
and provide two important benefits:
Keep a consistent and current local cache.
Any client will query local cache before trying to connect to and query the directory. Persistent
searches provide the local cache necessary to improve performance for these clients.
Automatically initiate directory actions.
T he persistent cache can be automatically updated as entries are modified, and the persistent
search results can display what kind of modification was performed on the entry. Another application
270
Chapter 8. Finding D irectory Entries
can use that output to update entries automatically, such as automatically creating an email account
on a mail server for new users or generating a unique user ID number.
T here are some performance considerations when running persistent searches, as well:
T he ldapsearch does not send a notification when the client disconnects, and the change
notifications are not sent for any changes made while the search is disconnected. T his means that
the client's cache will not be updated if it is ever disconnected and there is no good way to update
the cache with any new, modified, or deleted entries that were changed while it was disconnected.
An attacker could open a large number of persistent searches to launch a denial of service attack.
A persistent search requires leaving open a T CP connection between the Directory Server and client.
T his should only be done if the server is configured to allow a lot of client connections and has a way
to close idle connections.
Each persistent search runs as a separate thread inside the Directory Server.
8.5.1. An Overview of Persistent Searches
T he purpose of a persistent search is to provide a continuous list of changes to the directory entries as
well as the complete entries themselves, something like a hybrid search and changelog. T herefore, the
search command must specify what entries to return (the search parameters) and what changes cause
an entry to be returned (entry change parameters).
T he persistent search command sets the change information in three aspects:
Whether to return all of the currently matching entries and then update the cache with changed
entries or whether to only send changed entries which match the search parameters. T his is the
changesOnly setting.
What kinds of changes, such as adds or modifies, to entries are returned. T his is the changetype
setting.
Whether to display within the entry what kind of modification was performed. T his is the
entryChanges setting.
T he persistent search command, then, has the following format:
/usr/lib64/mozldap/ldapsearch -r -C PS:changetype:changesOnly:entryChanges -b
baseDN -s scope -D bindDN -w password -p port -h host (filter)
A persistent ldapsearch takes the normal ldapsearch options, such as the server connection
information and bind credentials.
Additionally, the -C argument and the PS: signal that this is a persistent search. T he -r argument is a
recommended ldapsearch argument that prints the entries immediately to the screen as soon as they
enter the buffer. T his prevents the search command from buffering its output, which ultimately makes the
search command appear to hang.
T he term changetype:changesOnly:entryChanges is what defines the entry change parameters.
T he only required term is the changetype, which sets what kinds of changes are returned in the
persistent search. T he possible values are add, m odify, m odrdn (for modrdn operations),
delete, and any.
changesOnly is optional. If this value is true (1), then no entries are returned in the initial search, and
entries are only returned as they are modified. If it is false (0), then all matching entries are returned
in the initial search, as well as sending updated entries. T he default is true.
entryChanges is also optional. If this value is true (1), then the type of modification is included in a
271
Red Hat D irectory Server 8.2 Administration Guide
changetype line in the returned entry. For example:
dn: uid=scarter,ou=People,dc=example,dc=com
persistentSearch-changetype: modify
If the value is false (0), then only the entry is returned, without the type of change being noted. T he
default is true.
8.5.2. Running a Persistent Search
T o run a persistent search, simply run an ldapsearch with the -C and -r arguments and the change
parameters for the search. If no bind credentials are given, the Directory Server is accessed by the
anonymous user.
/usr/lib64/mozldap/ldapsearch -D "cn=directory manager" -w secret -p 389 -h
server.example.com -r -C PS:any:1:0 -b dc=example,dc=com objectclass=person
T he search, if the changesOnly setting is false (0), will return every matching entry in the directory, as
with a normal ldapsearch. T hen, as entries are modified, the persistent search returns the newly
updated entry with, optionally, the type of operation which updated it (changeType).
version: 1
dn: uid=scarter,ou=People,dc=example,dc=com
persistentSearch-changetype: modify
title: manager
mail: [email protected]
description: test
uid: scarter
givenname: scott
objectClass: top
objectClass: person
objectClass: organizationalPerson
objectClass: inetorgperson
sn: carter
cn: scott carter
In the access logs, a persistent search is identifies with the tag options=persistent.
[12/Jan/2009:12:51:54 -0500] conn=19636710736396323 op=0 SRCH
base="dc=example,dc=com" scope=2 filter="(objectClass=person)" attrs=ALL
options=persistent
8.6. Performing Dereferencing Searches
A dereferencing search is a quick way to trackback over cross-references in an entry and return
information about the referenced entry. For example, a group entry contains references to its member's
user entries. A dereferencing search for the group entry can then return information about the members
— such as their locations, email addresses, or managers — along with the search results for the group.
Dereferencing simplifies many client operations. Cross-links show relationships between entries. Some
operations may require getting a list of cross-links from one entry and then performing a series of
subsequent searches to get information from each entry on the list. Dereferencing allows those
sequences of searches to be consolidated into a single search.
272
Chapter 8. Finding D irectory Entries
IMPORTANT
T he dereferencing searches are not done using MozLDAP command-line tools. T he server
supports dereferencing search operations; however the client tools with Red Hat Directory Server
do not. T herefore, dereferencing operations must be done using OpenLDAP command-line tools
version 2.4.18 or later or other clients which support dereferencing searches.
T o download the appropriate OpenLDAP tools, follow the directions at
http://www.openldap.org/faq/data/cache/897.html.
T he format of the dereference arguments is:
-E 'deref=deref_attribute:list_of_attributes'
T he deref_attribute is the attribute in the search target that contains the reference. T his can be any
attribute which has a DN for its value, such as m em ber or m anager.
NOTE
Not only must the value of the deref_attribute be a DN, but the actual defined syntax for the
attribute must be DN syntax (1.3.6.1.4 .1.14 66.115.121.1.12).
T he list_of_attributes is one or more attributes in the referenced entry which will be returned along with
the primary search results. Multiple attributes can be separated by commas, like l,m ail,cn.
Figure 8.3. Simple Dereferencing Search Command
T he requested dereferenced information requested in the search argument is returned with the rest of
the search results. For example, this dereferencing search tells the server to use the member attribute in
the search target entry (the Engineers group) as the deref_attribute. It then returns the locality attribute
273
Red Hat D irectory Server 8.2 Administration Guide
for each member.
ldapsearch -x -D "cn=Directory Manager" -w secret -b
"cn=Example,ou=Groups,dc=example,dc=com" -E 'deref=member:mail,cn'
"(objectclass=*)"
# Engineers, Groups, example.com
dn: cn=Engineers,ou=Groups,dc=example,dc=com
control: 1.3.6.1.4.1.4203.666.5.16 false MIQAAADNMIQAAAA1BAZtZW1iZXIEK2NuPURld
mVsb3BlcnMsIG91PUdyb3VwcywgZGM9ZXhhbXBsZSxkYz1jb20whAAAADIEBm1lbWJlcgQoY249VG
VzdGVycywgb3U9R3JvdXBzLCBkYz1leGFtcGxlLGRjPWNvbTCEAAAAVAQGbWVtYmVyBCp1aWQ9ZW5
nLCBvdT1lbmdpbmVlcmluZywgZGM9ZXhhbXBsZSxkYz1jb22ghAAAABowhAAAABQEAWwxhAAAAAsE
CUNhbWJyaWRnZQ==
# member: <[email protected]><cn=John
Smith>;uid=jsmith,ou=people,dc=example,dc=com
objectClass: top
objectClass: inetuser
objectClass: groupofnames
cn: Engineers
member: uid=jsmith,ou=people,dc=example,dc=com
8.7. Using Simple Paged Results
Search results can be very large, and a part of processing the results is organizing the results. One
method of doing this is using a server-side sort, which lists the results according to a sort order for an
attribute. Another method is using simple paged results, a control that breaks the results into pages of a
certain length.
T he simple paged results control sets the number of entries to display at a time. T he results can be
scrolled through, a page at a time, which makes the results easier to digest. T he full behavior of the
control is described in RFC 2696.
Simple paged results is implemented as an LDAP control extension for the Directory Server. Its OID is
1.2.84 0.113556.1.4 .319.
T he paging itself is done in the server frontend. T he search progresses as normal in the backend
database, going through resource limits like the look-through and time limits as it pulls candidate entries.
T he backend database then forwards the list of entry IDs to the server frontend, and the frontend
manipulates the results so that the appropriate page sizes are returned.
Figure 8.4 . Simple Paged Results Operation
274
Chapter 8. Finding D irectory Entries
IMPORTANT
Simple paged result searches are not done using MozLDAP command-line tools. T he server
supports simple paged results; however the client tools with Red Hat Directory Server do not.
T herefore, simple paged results operations must be done using OpenLDAP command-line tools
version 2.4.18 or later or other clients which support simple paged results, such as Perl
Net::LDAP.
T o download the appropriate OpenLDAP tools, follow the directions at
http://www.openldap.org/faq/data/cache/897.html.
T he format of the simple paged result search option with OpenLDAP's ldapsearch is:
-E pg=size
T he size value is the page size, or the number of entries to include per page. For example, using
OpenLDAP's ldapsearch:
ldapsearch -x -D "cn=Directory Manager" -w secret -b
"ou=Engineers,ou=People,dc=example,dc=com" -E pg=3 "(objectclass=*)" cn
dn: uid=jsmith,ou=Engineers,ou=People,dc=example,dc=com
cn: John Smith
dn: uid=bjensen,ou=Engineers,ou=People,dc=example,dc=com
cn: Barbara Jensen
dn: uid=hmartin,ou=Engineers,ou=People,dc=example,dc=com
cn: Henry Martin
Results are sorted.
next page size (3): 5
T he tag at the end shows the configured page size (the number in parentheses) from the search. After
the colon, one enters the page size for the next page, so entering 5 as shown would open the next page
of results with five entries.
Simple paged results can be used together with server-side sorting. Server-side sorting is a control
which performs the sort process on the server rather than in a client; this is usually done for a search
which uses a particular matching rule. (T his behavior is defined in RFC 2891.) T he OpenLDAP client
tools do not support server-side sort with the simple paged results control, but other LDAP utilities such
as Perl Net::LDAP do support both.
VLV indexes are similar to simple paged results in that they also return a usable browsing list of results.
T he main difference is in how that list is generated. Simple paged results are calculated per search,
while VLV indexes are a permanent list. Overall, VLV indexes are faster for searches, but do require
some server-side configuration and overhead for the server to maintain.
275
Red Hat D irectory Server 8.2 Administration Guide
NOTE
Simple paged results and VLV indexes cannot be used on the same search. Simple paged results
would attempt to manipulate the VLV index, which is already a browsin index. If the control is
passed for a search using a VLV index, then the server returns an UNWILLING_T O_PERFORM
error.
For more information on VLV indexes, see Section 7.4, “Creating Browsing (VLV) Indexes”.
276
Chapter 9. Managing Replication
Chapter 9. Managing Replication
Replication is the mechanism by which directory data is automatically copied from one Red Hat Directory
Server instance to another; it is an important mechanism for extending the directory service beyond a
single server configuration. T his chapter describes the tasks to be performed on the master and
consumer servers to set up single-master replication, multi-master replication, and cascading replication.
9.1. Replication Overview
Replication is the mechanism by which directory data is automatically copied from one Directory Server
to another. Updates of any kind — entry additions, modifications, or even deletions — are automatically
mirrored to other Directory Servers using replication.
Section 9.1.1, “What Directory Units Are Replicated”
Section 9.1.2, “Read-Write and Read-Only Replicas”
Section 9.1.3, “Suppliers and Consumers”
Section 9.1.4, “Changelog”
Section 9.1.5, “Replication Identity”
Section 9.1.6, “Replication Agreement”
Section 9.1.8, “Compatibility with Earlier Versions of Directory Server”
9.1.1. What Directory Units Are Replicated
T he smallest unit of of the directory which can be replicated is a database. T his means that one can
replicate an entire database but not a subtree within a database. T herefore, when creating the directory
tree, consider any replication plans as part of determining how to distribute information.
Replication also requires that one database correspond to one suffix. T his means that a suffix (or
namespace) that is distributed over two or more databases using custom distribution logic cannot be
replicated. For more information on this topic, see Section 2.2, “Creating and Maintaining Databases”.
9.1.2. Read-Write and Read-Only Replicas
A database that participates in replication is called a replica. T here are two kinds of replicas: read-write
or read-only. A read-write replica contains master copies of directory information and can be updated. A
read-only replica services read, search, and compare requests, but refers all update operations to readwrite replicas. A server can hold any number of read-only or read-write replicas.
9.1.3. Suppliers and Consumers
A server that holds a replica that is copied to a replica on a different server is called a supplier for that
replica. A server that holds a replica that is copied from a different server is called a consumer for that
replica. Generally, the replica on the supplier server is a read-write replica, and the one on the consumer
server is a read-only replica, with two exceptions:
In the case of cascading replication, the hub server holds a read-only replica that it supplies to
consumers. Section 9.2.3, “Cascading Replication” has more information.
In the case of multi-master replication, the masters are both suppliers and consumers for the same
information. For more information, see Section 9.2.2, “Multi-Master Replication”.
Replication is always initiated by the supplier server, never by the consumer (supplier-initiated
replication). Supplier-initiated replication allows a supplier server to be configured to push data to
multiple consumer servers.
277
Red Hat D irectory Server 8.2 Administration Guide
9.1.4. Changelog
Every supplier server maintains a changelog, a record of all changes that a supplier or hub needs to
send to its consumers. A changelog is a special kind of database that describes the modifications that
have occurred on a replica. T he supplier server then replays these modifications to the replicas stored
on consumer servers or to other suppliers, in the case of multi-master replication.
When an entry is modified, a change record describing the LDAP operation that was performed is
recorded in the changelog.
In Directory Server, the changelog is only intended for internal use by the server. For other applications
to read the changelog, use the Retro Changelog Plug-in, as described in Section 9.16, “Using the Retro
Changelog Plug-in”.
9.1.5. Replication Identity
When replication occurs between two servers, the replication process uses a special entry, called the
replication manager entry, to identify replication protocol exchanges and to control access to the
directory data. T he replication manager entry, or any entry used during replication, must meet the
following criteria:
It is created on the consumer server (or hub) and not on the supplier server.
Create this entry on every server that receives updates from another server, meaning on every hub
or dedicated consumer.
When a replica is configured as a consumer or hub (a replica which receives updates from another
server), this entry must be specified as the one authorized to perform replication updates.
T he replication agreement is created on the supplier server, the DN of this entry must be specified in
the replication agreement.
T he supplier bind DN entry must not be part of the replicated database for security reasons.
T his entry, with its special user profile, bypasses all access control rules defined on the consumer
server for the database involved in that replication agreement.
NOTE
In the Directory Server Console, this replication manager entry is referred to as the supplier bind
DN, which may be misleading because the entry does not actually exist on the supplier server. It
is called the supplier bind DN because it is the entry which the supplier uses to bind to the
consumer. T his entry actually exists, then, on the consumer.
For more information on creating the replication manager entry, see Section 9.3, “Creating the Supplier
Bind DN Entry”.
9.1.6. Replication Agreement
Directory Servers use replication agreements to define their replication configuration. A replication
agreement describes replication between one supplier and one consumer only. T he agreement is
configured on the supplier server and must specify all required replication information:
T he database to be replicated.
T he consumer server to which the data is pushed.
T he days and times during which replication can occur.
T he DN and credentials that the supplier server must use to bind (the replication manager entry or
supplier bind DN).
278
Chapter 9. Managing Replication
How the connection is secured (SSL, client authentication).
Any attributes that will not be replicated (fractional replication).
9.1.7. Replicating a Subset of Attributes with Fractional Replication
Fractional replication sets a specific subset of attributes that will not be transmitted from a supplier to the
consumer (or another supplier). Administrators can therefore replicate a database without replicating all
the information that it contains or all of the information in every entry.
Fractional replication is enabled and configured per replication agreement. Excluding attributes from
replication is applied equally to all entries within the replication agreement's scope.
As far as the consumer server is concerned, the excluded attributes always have no value. T herefore, a
client performing a search against the consumer server will never see the excluded attributes. Similarly,
should it perform a search that specifies those attributes in its filter, no entries will match.
9.1.8. Compatibility with Earlier Versions of Directory Server
T he replication mechanism in Directory Server 8.2 is different from the mechanism used in 4.x of
Directory Server. Compatibility is provided through two Directory Server plug-ins:
Legacy Replication Plug-in. T he Legacy Replication Plug-in makes a Directory Server 8.2 instance
behave as a 4.x Directory Server in a consumer role. For information on how to implement legacy
replication using this plug-in, see Section 9.15, “Replication with Earlier Releases”.
Retro Changelog Plug-in. T he Retro Changelog Plug-in can be used for a Directory Server supplier
to maintain a 4.x-style changelog. T his is sometimes necessary for legacy applications that have a
dependency on the Directory Server 4.x changelog format because they read information from the
changelog. For more information on the Retro Changelog Plug-in, see Section 9.16, “Using the Retro
Changelog Plug-in”.
9.2. Replication Scenarios
Section 9.2.1, “Single-Master Replication”
Section 9.2.2, “Multi-Master Replication”
Section 9.2.3, “Cascading Replication”
T hese basic strategies can be combined in a variety of ways to create the best replication environment.
NOTE
Whatever replication scenario is implemented, consider schema replication. T o avoid conflict
resolution loops, the Referential Integrity Plug-in should only be enabled on one supplier replica in
a multi-master replication environment. T he plug-in is off by default.
9.2.1. Single-Master Replication
In the simplest replication scenario, the master copy of directory data is held in a single read-write
replica on one server called the supplier server. T he supplier server also maintains changelog for this
replica. On another server, called the consumer server, there can be multiple read-only replicas. Such
scenarios are called single-master configurations. Figure 9.1, “Single-Master Replication” shows an
example of single-master replication.
279
Red Hat D irectory Server 8.2 Administration Guide
Figure 9.1. Single-Master Replication
In this particular configuration, the ou=people,dc=exam ple,dc=com suffix receives a large number
of search requests. T herefore, to distribute the load, this tree, which is mastered on Server A, is
replicated to two read-only replicas located on Server B and Server C.
For information on setting up a single-master replication environment, see Section 9.4, “Configuring
Single-Master Replication”.
9.2.2. Multi-Master Replication
Directory Server also supports complex replication scenarios in which the same suffix (database) can be
mastered on many servers. T his suffix is held in a read-write replica on each server. T his means that
each server maintains a changelog for the read-write replica.
Multi-master replication in Directory Server supports up to four masters, an unlimited number of hub
suppliers, and an unlimited number of consumer servers. Each consumer server holds a read-only
replica. T he consumers can receive updates from any or all the suppliers. T he consumers also have
referrals defined for all the suppliers to forward any update requests that the consumers receive. Such
scenarios are called multi-master configurations.
Figure 9.2, “Multi-Master Replication (T wo Masters)” shows an example of multi-master replication
scenario with two supplier servers and two consumer servers.
280
Chapter 9. Managing Replication
Figure 9.2. Multi-Master Replication (T wo Masters)
Figure 9.3, “Multi-Master Replication (Four Masters)” shows a sample of multi-master replication
scenario with four supplier servers and eight consumer servers. In this sample setup, each supplier
server is configured with ten replication agreements to feed data to two other supplier servers and all
eight consumer servers.
281
Red Hat D irectory Server 8.2 Administration Guide
Figure 9.3. Multi-Master Replication (Four Masters)
Multi-master configurations have the following advantages:
Automatic write failover when one supplier is inaccessible.
Updates are made on a local supplier in a geographically distributed environment.
NOTE
T he speed that replication proceeds depends on the speed of the network. Plan changes and
directory configuration accordingly, and realize that changes to one directory may not be quickly
replicated to other directories over slow links, such as wide-area networks, in geographicallydistributed environments.
For the procedure to set up multi-master replication, see Section 9.5, “Configuring Multi-Master
Replication”.
9.2.3. Cascading Replication
In a cascading replication scenario, one server, a hub, acts both as a consumer and a supplier. It holds a
282
Chapter 9. Managing Replication
read-only replica and maintains a changelog, so it receives updates from the supplier server that holds
the master copy of the data and, in turn, supplies those updates to the consumer. Cascading replication
is very useful for balancing heavy traffic loads or to keep master servers based locally in geographicallydistributed environments.
Figure 9.4, “Cascading Replication” shows an example of a simple cascading replication scenario,
though it is possible to create more complex scenarios with several hub servers.
Figure 9.4 . Cascading Replication
For information on setting up cascading replication, see Section 9.6, “Configuring Cascading Replication”.
NOTE
Multi-master and cascading replication can be combined. For example, in the multi-master
scenario illustrated in Figure 9.2, “Multi-Master Replication (T wo Masters)”, Server C and Server
D could be hub servers that would replicate to any number of consumer servers.
9.3. Creating the Supplier Bind DN Entry
A critical part of setting up replication is to create the entry, called the replication manager or supplier
bind DN entry, that the suppliers use to bind to the consumer servers to perform replication updates.
T he supplier bind DN must meet the following criteria:
283
Red Hat D irectory Server 8.2 Administration Guide
It must be unique.
It must be created on the consumer server (or hub) and not on the supplier server.
It must correspond to an actual entry on the consumer server.
It must be created on every server that receives updates from another server.
It must not be part of the replicated database for security reasons.
It must be defined in the replication agreement on the supplier server.
It must have an idle timeout period set to a high enough limit to allow the initialization process for
large databases to complete. Using the nsIdleTimeOut operational attribute allows the replication
manager entry to override the global nsslapd-idletimeout setting.
For example, the entry cn=Replication Manager,cn=config can be created under the
cn=config tree on the consumer server. T his would be the supplier bind DN that all supplier servers
would use to bind to the consumer to perform replication operations.
NOTE
Avoid creating simple entries under the cn=config entry in the dse.ldif file. T he
cn=cn=config entry in the simple, flat dse.ldif configuration file is not stored in the same
highly scalable database as regular entries. As a result, if many entries, and particularly entries
that are likely to be updated frequently, are stored under cn=config, performance will suffer.
However, although Red Hat recommends not storing simple user entries under cn=config for
performance reasons, it can be useful to store special user entries such as the Directory
Manager entry or replication manager (supplier bind DN) entry under cn=config since this
centralizes configuration information.
On each server that acts as a consumer in replication agreements, create a special entry that the
supplier will use to bind to the consumers. Make sure to create the entry with the attributes required by
the authentication method specified in the replication agreement.
1. Stop the Directory Server. If the server is not stopped, the changes to the dse.ldif file will not
be saved. See Section 1.4, “Starting and Stopping Servers” for more information on stopping the
server.
2. Create a new entry, such as cn=replication m anager,cn=config, in the dse.ldif file.
3. Specify a userPassword attribute-value pair.
4. Set an nsIdleTimeout period that gives the replication user a long enough time limit to allow
replication initialization on large databases to complete.
5. If password expiration policy is enabled or ever will be enabled, disable it on the replication
manager entry to prevent replication from failing due to passwords expiring. T o disable the
password expiration policy on the userPassword attribute, add the passwordExpirationTime
attribute with a value of 203801190314 07Z, which means that the password will never expire.
6. Restart the Directory Server. See Section 1.4, “Starting and Stopping Servers” for more information
on starting the server.
T he final entry should resemble Example 9.1, “Example Supplier Bind DN Entry”.
284
Chapter 9. Managing Replication
Example 9.1. Example Supplier Bind DN Entry
dn: cn=replication manager,cn=config
objectClass: inetorgperson
objectClass: person
objectClass: top
cn: replication manager
sn: RM
userPassword: password
passwordExpirationTime: 20380119031407Z
nsIdleTimeout: 0
When configuring a replica as a consumer, use the DN of this entry to define the supplier bind DN.
9.4. Configuring Single-Master Replication
T o set up single-master replication such as the configuration shown in Figure 9.1, “Single-Master
Replication”, between supplier Server A, which holds a read-write replica, and the two consumers Server
B and Server C, which each hold a read-only replica, there are three major steps:
Section 9.4.1, “Configuring the Read-Write Replica on the Supplier Server”
Section 9.4.2, “Configuring the Read-Only Replica on the Consumer”
Section 9.4.3, “Creating the Replication Agreement”
9.4.1. Configuring the Read-Write Replica on the Supplier Server
1. Specify the supplier settings for the server.
a. In the Directory Server Console, select the Configuration tab.
b. In the navigation tree, select the Replication folder.
c. In the right-hand side of the window, select the Supplier Settings tab.
d. Check the Enable Changelog checkbox.
285
Red Hat D irectory Server 8.2 Administration Guide
T his activates all of the fields in the pane below that were previously grayed out.
e. Specify a changelog by clicking the Use default button, or click the Browse button to
display a file selector.
f. Set the changelog parameters for the number and age of the log files.
Clear the unlimited checkboxes to specify different values.
g. Click Save.
2. Specify the replication settings required for a read-write replica.
a. In the navigation tree on the Configuration tab, expand the Replication node, and
highlight the database to replicate.
T he Replica Settings tab opens in the right-hand side of the window.
b. Check the Enable Replica checkbox.
c. In the Replica Role section, select the Single Master radio button.
d. In the Com m on Settings section, specify a Replica ID, which is an integer between 1
and 65534 , inclusive.
T he replica ID must be unique for a given suffix, different from any other ID used for readwrite replicas on this server and on other servers.
e. In the Com m on Settings section, specify a purge delay in the Purge delay field.
T he purge delay is how often the state information stored for the replicated entries is
deleted.
f. Click Save.
9.4.2. Configuring the Read-Only Replica on the Consumer
1. Create the database for the read-only replica if it does not exist. See Section 2.1.1, “Creating
Suffixes” for instructions on creating suffixes.
2. Create the entry for the supplier bind DN on the consumer server if it does not exist. T he supplier
bind DN is the special entry that the supplier will use to bind to the consumer. T his is described in
Section 9.3, “Creating the Supplier Bind DN Entry”.
3. Specify the replication settings required for a read-only replica.
286
Chapter 9. Managing Replication
3. Specify the replication settings required for a read-only replica.
a. In the Directory Server Console, select the Configuration tab.
b. In the navigation tree, expand the Replication folder, and highlight the replica database.
T he Replica Settings tab for that database opens in the right-hand side of the window.
c. Check the Enable Replica checkbox.
d. In the Replica Role section, select the Dedicated Consum er radio button.
e. In the Com m on Settings section, specify a purge delay in the Purge delay field.
T his option indicates how often the state information stored for the replicated entries is
purged.
f. In the Update Settings section, specify the bind DN that the supplier will use to bind to
the replica. Enter the supplier bind DN in the Enter a new Supplier DN field, and click
Add. T he supplier bind DN appears in the Current Supplier DNs list.
287
Red Hat D irectory Server 8.2 Administration Guide
T he supplier bind DN should be the entry created in step 2. T he supplier bind DN is a
privileged user because it is not subject to access control.
NOTE
T here can be multiple supplier bind DNs per consumer but only one supplier DN per
replication agreement.
g. Specify the URL for any supplier servers to which to refer updates.
By default, all updates are first referred to the supplier servers that are specified here. If no
suppliers are set here, updates are referred to the supplier servers that have a replication
agreement that includes the current replica.
Automatic referrals assume that clients bind over a regular connection; this has a URL in
the form ldap://hostname:port. For clients to bind to the supplier using SSL, use this field
to specify a referral of the form ldaps://hostname:port, where the s in ldaps indicates a
secure connection.
4. Click Save.
Repeat these steps for every consumer server in the replication configuration.
9.4.3. Creating the Replication Agreement
Create one replication agreement for each read-only replica. For example, in the scenario illustrated in
Figure 9.1, “Single-Master Replication”, Server A has two replication agreements, one for Server B and
one for Server C.
1. In the navigation tree of the Configuration tab, right-click the database to replicate, and select
New Replication Agreem ent.
288
Chapter 9. Managing Replication
Alternatively, highlight the database, and select New Replication Agreem ent from the
Object menu to start the Replication Agreement Wizard.
2. In the first screen, fill in a name and description for the replication agreement, and hit Next.
3. In the Source and Destination screen, fill in the URL for the consumer and the supplier bind
DN and password on that consumer. If the target server is not available, hit in other to fill in the
information manually.
Unless there is more than one instance of Directory Server configured, by default, there are no
consumers available in the drop-down menu.
T he port listed is the non-SSL port, even if the Directory Server instance is configured to run
289
Red Hat D irectory Server 8.2 Administration Guide
T he port listed is the non-SSL port, even if the Directory Server instance is configured to run
over SSL. T his port number is used only for identification of the Directory Server instance in
the Console; it does not specify the actual port number or protocol that is used for replication.
If SSL is enabled on the servers, it is possible to select the Using encrypted SSL
connection radio button for SSL client authentication. Otherwise, fill in the supplier bind DN
and password.
NOTE
If attribute encryption is enabled, a secure connection must be used for the encrypted
attributes to be replicated.
4. Select the connection type. T here are three options:
Use LDAP. T his sets a standard, unencrypted connection.
Use TLS/SSL. T his uses a secure connection over the server's secure LDAPS port, such as
636. T his setting is required to use T LS/SSL.
Use Start TLS. T his uses Start T LS to establish a secure connection over the server's
standard port.
NOTE
If secure binds are required for simple password authentication (Section 13.5.1, “Requiring
Secure Binds”), then any replication operations will fail unless they occur over a secure
connection. Using a secure connection (SSL/T LS and Start T LS connections or SASL
authentication) is recommended, anyway.
5. Select the appropriate authentication method and supply the required information. T his gives the
information that the supplier uses to authenticate and bind to the consumer server to send
updates.
Simple means that the server connects over the standard port with no encryption. T he only
required information is the bind DN and password for the Replication Manager (which must
exist on the consumer server).
Server TLS/SSL Certificate uses the supplier's SSL certificate to authenticate to the consumer
server. A certificate must be installed on the supplier for certificate-based authentication, and
the consumer server must have certificate mapping configured so that it can map the subject
DN in the supplier's certificate to its Replication Manager entry.
Configuring SSL and certificate mapping is described in Section 14.2, “Using T LS/SSL”.
SASL/DIGEST-MD5, like simple authentication, requires only the bind DN and password to
authenticate. T his can run over a standard or SSL/T LS connection.
SASL/GSSAPI requires the supplier server to have a Kerberos keytab (as in Section 14.3.4.2,
“About the KDC Server and Keytabs”), and the consumer server to have a SASL mapping to
map the supplier's principal to the real replication manager entry (as in Section 14.3.5.1,
“Configuring SASL Identity Mapping from the Console”).
6. Fractional replication controls which entry attributes are replicated between servers. By default, all
attributes are replicated. T o select attributes that will not be replicated to the consumer, check the
Enable Fractional Replication checkbox. T hen, highlight the attribute (or attributes) in
the Included column on the right, and click Rem ove. All attributes that will not be replicated are
listed in the Excluded column on the left, as well as in the summary the replication agreement is
complete.
290
Chapter 9. Managing Replication
7. Set the schedule for when replication runs. By default, replication runs continually.
291
Red Hat D irectory Server 8.2 Administration Guide
NOTE
T he replication schedule cannot cross midnight (0000). So, it is possible to set a schedule
that begins at 0001 and ends at 2359 on the same day, but it is not possible to set one
that begins at 2359 on one day and ends at 0001 on the next.
Hit Next.
8. Set when the consumer is initialized. Initializing a consumer manually copies all data over from the
supplier to the consumer. T he default is to create an initialization file (an LDIF of all supplier data)
so that the consumer can be initialized later. It is also possible to initialize the consumer as soon
as the replication agreement is completed or not at all. For information on initializing consumers,
see Section 9.10, “Initializing Consumers”.
NOTE
Replication will not begin until the consumer is initialized.
Hit Next.
9. T he final screen shows the settings for the replication agreement, as it will be included in the
dse.ldif file. Hit Done to save the agreement.
292
Chapter 9. Managing Replication
T he replication agreement is set up.
NOTE
After creating a replication agreement, the connection type (SSL or non-SSL) cannot be changed
because LDAP and LDAPS connections use different ports. T o change the connection type, recreate the replication agreement.
9.5. Configuring Multi-Master Replication
In a multi-master configuration, many suppliers can accept updates, synchronize with each other, and
update all consumers. T he consumers can send referrals for updates to all masters.
Directory Server supports four-way multi-master replication, meaning that there can be up to four
masters (and an unlimited number of hub suppliers) in a single replication scenario. Directory Server
allows an unlimited number of consumers.
T o set up multi-master replication, set up all of the consumers first, then set up the suppliers, and last,
initialize all of the databases.
Section 9.5.1, “Configuring the Read-Write Replicas on the Supplier Servers”
Section 9.5.2, “Configuring the Read-Only Replicas on the Consumer Servers”
Section 9.5.3, “Setting up the Replication Agreements”
Section 9.5.4, “Preventing Monopolization of the Consumer in Multi-Master Replication”
293
Red Hat D irectory Server 8.2 Administration Guide
NOTE
More than 10 databases running with replication or more than on a supplier can cause
performance degradation. T o support that many consumers, introduce hub replicas between the
suppliers and consumers. See Section 9.6, “Configuring Cascading Replication”.
9.5.1. Configuring the Read-Write Replicas on the Supplier Servers
Set up each supplier server. T he first supplier configured should be used to initialize the other suppliers
in the multi-master replication environment.
1. Specify the supplier settings for the server.
a. In the Directory Server Console, select the Configuration tab.
b. In the navigation tree, select the Replication folder.
c. In the right-hand side of the window, select the Supplier Settings tab.
d. Check the Enable Changelog checkbox.
T his activates all of the fields in the pane below that were previously grayed out.
e. Specify a changelog by clicking the Use default button, or click the Browse button to
display a file selector.
f. Set the changelog parameters for the number and age of the log files.
Clear the unlimited checkboxes to specify different values.
g. Click Save.
2. Create the entry for the supplier bind DN on the consumer server if it does not exist. T his is the
special entry that the other suppliers will use to bind to this supplier, as in other supplierconsumer relationships. T his is described in Section 9.3, “Creating the Supplier Bind DN Entry”.
294
Chapter 9. Managing Replication
NOTE
For multi-master replication, it is necessary to create this supplier bind DN on the supplier
servers as well as the consumers because the suppliers act as both consumer and
supplier to the other supplier servers.
3. Specify the replication settings for the multi-mastered read-write replica.
a. In the Directory Server Console, select the Configuration tab.
b. In the navigation tree, expand the Replication folder, and highlight the replica database.
T he Replica Settings tab for that database opens in the right-hand side of the window.
c. Check the Enable Replica checkbox.
d. In the Replica Role section, select the Multiple Master radio button.
e. In the Com m on Settings section, specify a Replica ID, which is an integer between 1
and 65534 , inclusive.
T he replica ID must be unique for a given suffix, different from any other ID used for readwrite replicas on this server and on other servers.
f. In the Com m on Settings section, specify a purge delay in the Purge delay field.
T he purge delay is how often the state information stored for the replicated entries is
deleted.
g. In the Update Settings section, specify the bind DN that the supplier will use to bind to
the replica. Enter the supplier bind DN in the Enter a new Supplier DN field, and click
Add. T he supplier bind DN appears in the Current Supplier DNs list.
295
Red Hat D irectory Server 8.2 Administration Guide
T he supplier bind DN should be the entry created in step 2. T he supplier bind DN is a
privileged user because it is not subject to access control in the replicated database.
NOTE
T here can be multiple supplier bind DNs per consumer but only one supplier DN per
replication agreement.
h. Specify the URL for any supplier servers to which to refer updates, such as the other
suppliers in the multi-master replication set. Only specify the URL for the supplier server.
For clients to bind using SSL, specify a URL beginning with ldaps://.
i. Click Save.
9.5.2. Configuring the Read-Only Replicas on the Consumer Servers
First, configure every consumer before creating any replication agreements.
1. Create the database for the read-only replica if it does not exist. See Section 2.1.1, “Creating
Suffixes” for instructions on creating suffixes.
2. Create the entry for the supplier bind DN on the consumer server if it does not exist. T he supplier
bind DN is the special entry that the supplier will use to bind to the consumer. T his is described in
Section 9.3, “Creating the Supplier Bind DN Entry”.
3. Specify the replication settings required for a read-only replica.
a. In the Directory Server Console, select the Configuration tab.
b. In the navigation tree, expand the Replication folder, and highlight the replica database.
T he Replica Settings tab for that database opens in the right-hand side of the window.
296
Chapter 9. Managing Replication
c. Check the Enable Replica checkbox.
d. In the Replica Role section, select the Dedicated Consum er radio button.
e. In the Com m on Settings section, specify a purge delay in the Purge delay field.
T his option indicates how often the state information stored for the replicated entries is
purged.
f. In the Update Settings section, specify the bind DN that the supplier will use to bind to
the replica. Enter the supplier bind DN in the Enter a new Supplier DN field, and click
Add. T he supplier bind DN appears in the Current Supplier DNs list.
T he supplier bind DN should be the entry created in step 2. T he supplier bind DN is a
privileged user because it is not subject to access control in the replicated database.
297
Red Hat D irectory Server 8.2 Administration Guide
NOTE
T here can be multiple supplier bind DNs per consumer but only one supplier DN per
replication agreement.
g. Specify the URL for any supplier servers to which to refer updates.
By default, all updates are first referred to the supplier servers that are specified here. If no
suppliers are set here, updates are referred to the supplier servers that have a replication
agreement that includes the current replica.
Automatic referrals assume that clients bind over a regular connection; this has a URL in
the form ldap://hostname:port. For clients to bind to the supplier using SSL, use this field
to specify a referral of the form ldaps://hostname:port, where the s in ldaps indicates a
secure connection.
4. Click Save.
Repeat these steps for every consumer server in the replication configuration.
9.5.3. Setting up the Replication Agreements
NOTE
1. First set up replication agreements on a single supplier, the data master, between the other
multi-master suppliers, and initialize all of the other suppliers.
2. T hen create replication agreements for all other suppliers in the multi-master replication
set, but do not reinitialize any of the suppliers.
3. T hen create replication agreements for all of the consumers from the single data master,
and initialize the consumers.
4. T hen create replication agreements for all of the consumers from for all of the other
suppliers, but do not reinitialize any of the consumers.
1. In the navigation tree of the Configuration tab, right-click the database to replicate, and select
New Replication Agreem ent.
Alternatively, highlight the database, and select New Replication Agreem ent from the
Object menu to start the Replication Agreement Wizard.
2. In the first screen, fill in a name and description for the replication agreement, and hit Next.
3. In the Source and Destination screen, fill in the URL for the consumer and the supplier bind
298
Chapter 9. Managing Replication
DN and password on that consumer. If the target server is not available, hit in other to fill in the
information manually.
Unless there is more than one instance of Directory Server configured, by default, there are no
consumers available in the drop-down menu.
T he port listed is the non-SSL port, even if the Directory Server instance is configured to run
over SSL. T his port number is used only for identification of the Directory Server instance in
the Console; it does not specify the actual port number or protocol that is used for replication.
If SSL is enabled on the servers, it is possible to select the Using encrypted SSL
connection radio button for SSL client authentication. Otherwise, fill in the supplier bind DN
and password.
NOTE
If attribute encryption is enabled, a secure connection is required for the encrypted
attributes to be replicated.
4. Select the connection type. T here are three options:
Use LDAP. T his sets a standard, unencrypted connection.
Use TLS/SSL. T his uses a secure connection over the server's secure LDAPS port, such as
299
Red Hat D irectory Server 8.2 Administration Guide
636. T his setting is required to use T LS/SSL.
Use Start TLS. T his uses Start T LS to establish a secure connection over the server's
standard port.
NOTE
If secure binds are required for simple password authentication (Section 13.5.1, “Requiring
Secure Binds”), then any replication operations will fail unless they occur over a secure
connection. Using a secure connection (SSL/T LS and Start T LS connections or SASL
authentication) is recommended, anyway.
5. Select the appropriate authentication method and supply the required information. T his gives the
information that the supplier uses to authenticate and bind to the consumer server to send
updates.
Simple means that the server connects over the standard port with no encryption. T he only
required information is the bind DN and password for the Replication Manager (which must
exist on the consumer server).
Server TLS/SSL Certificate uses the supplier's SSL certificate to authenticate to the consumer
server. A certificate must be installed on the supplier for certificate-based authentication, and
the consumer server must have certificate mapping configured so that it can map the subject
DN in the supplier's certificate to its Replication Manager entry.
Configuring SSL and certificate mapping is described in Section 14.2, “Using T LS/SSL”.
SASL/DIGEST-MD5, like simple authentication, requires only the bind DN and password to
authenticate. T his can run over a standard or SSL/T LS connection.
SASL/GSSAPI requires the supplier server to have a Kerberos keytab (as in Section 14.3.4.2,
“About the KDC Server and Keytabs”), and the consumer server to have a SASL mapping to
map the supplier's principal to the real replication manager entry (as in Section 14.3.5.1,
“Configuring SASL Identity Mapping from the Console”).
6. Hit Next.
7. Fractional replication controls which entry attributes are replicated between servers. By default, all
attributes are replicated. T o select attributes that will not be replicated to the consumer, check the
Enable Fractional Replication checkbox. T hen, highlight the attribute (or attributes) in
the Included column on the right, and click Rem ove. All attributes that will not be replicated are
listed in the Excluded column on the left, as well as in the summary the replication agreement is
complete.
300
Chapter 9. Managing Replication
8. Set the schedule for when replication runs. By default, replication runs continually.
301
Red Hat D irectory Server 8.2 Administration Guide
NOTE
T he replication schedule cannot cross midnight (0000). So, it is possible to set a schedule
that begins at 0001 and ends at 2359 on the same day, but it is not possible to set one
that begins at 2359 on one day and ends at 0001 on the next.
Hit Next.
9. Set when the consumer is initialized. Initializing a consumer manually copies all data over from the
supplier to the consumer. T he default is to create an initialization file (an LDIF of all supplier data)
so that the consumer can be initialized later. It is also possible to initialize the consumer as soon
as the replication agreement is completed or not at all. For information on initializing consumers,
see Section 9.10, “Initializing Consumers”. For multi-master replication, consider the following:
Ensure one supplier has the complete set of data to replicate to the other suppliers. Use this
one supplier to initialize the replica on all other suppliers in the multi-master replication set.
Initialize the replicas on the consumer servers from any of the multi-master suppliers.
Do not try to reinitialize the servers when the replication agreements are set up. For example,
do not initialize server1 from server2 if server2 has already been initialized from server1. In
this case, select Do not initialize consum er.
NOTE
Replication will not begin until the consumer is initialized.
IMPORTANT
For multi-master replication, be sure that consumers are only initialized once, by one
supplier. When checking the replication status, be sure to check the replication agreement
entry, on the appropriate supplier, which was used to initialize the consumer.
302
Chapter 9. Managing Replication
Hit Next.
10. T he final screen shows the settings for the replication agreement, as it will be included in the
dse.ldif file. Hit Done to save the agreement.
T he replication agreement is set up.
NOTE
At the end of this procedure, all supplier servers will have mutual replication agreements, which
means that they can accept updates from each other.
NOTE
After creating a replication agreement, the connection type (SSL or non-SSL) cannot be changed
because LDAP and LDAPS connections use different ports. T o change the connection type, recreate the replication agreement.
9.5.4. Preventing Monopolization of the Consumer in Multi-Master Replication
One of the features of multi-master replication is that a supplier acquires exclusive access to the
consumer for the replicated area. During this time, other suppliers are locked out of direct contact with
the consumer. If a supplier attempts to acquire access while locked out, the consumer sends back a
busy response, and the supplier sleeps for several seconds before making another attempt. Normally,
this is all right; the supplier simply sends its update to another consumer while the first consumer is
locked and then send updates when the first consumer is free again.
303
Red Hat D irectory Server 8.2 Administration Guide
A problem can arise if the locking supplier is under a heavy update load or has a lot of pending updates
in the changelog. If the locking supplier finishes sending updates and then has more pending changes
to send, it will immediately attempt to reacquire the consumer and will most likely succeed, since the
other suppliers usually will be sleeping. T his can cause a single supplier to monopolize a consumer for
several hours or longer.
T wo attributes address this issue, nsds5ReplicaBusyWaitTime and
nsds5ReplicaSessionPauseTime.
nsds5ReplicaBusyWaitTime. T he nsds5ReplicaBusyWaitTime attribute sets the amount of time in
seconds a supplier should wait after a consumer sends back a busy response before making
another attempt to acquire access. T he default is 3 seconds.
nsds5ReplicaSessionPauseTime. T he nsds5ReplicaSessionPauseTime attribute sets the amount
of time in seconds a supplier should wait between update sessions. Set this interval so that it is at
least one second longer than the interval specified for nsds5ReplicaBusyWaitTime. Increase the
interval as needed until there is an acceptable distribution of consumer access among the suppliers.
T he default is 0.
T hese two attributes may be present in the nsds5ReplicationAgreement object class, which is used
to describe replication agreements. Set the attributes at any time by using changetype:m odify with
the replace operation. T he change takes effect for the next update session if one is already in
progress.
NOTE
If either attribute is set to a negative value, Directory Server sends the client a message and an
LDAP_UNWILLING_T O_PERFORM error code.
T he two attributes are designed so that the nsds5ReplicaSessionPauseTime interval will always be at
least one second longer than the interval specified for nsds5ReplicaBusyWaitTime. T he longer
interval gives waiting suppliers a better chance to gain consumer access before the previous supplier
can re-access the consumer.
If either attribute is specified but not both, nsds5ReplicaSessionPauseTime is set automatically to
1 second more than nsds5ReplicaBusyWaitTime.
If both attributes are specified, but nsds5ReplicaSessionPauseTime is less than or equal to
nsds5ReplicaBusyWaitTime, nsds5ReplicaSessionPauseTime is set automatically to 1 second
more than nsds5ReplicaBusyWaitTime.
If Directory Server has to automatically reset the value of nsds5ReplicaSessionPauseTime, the value
is changed internally only. T he change is not visible to clients, and it not saved to the configuration file.
From an external viewpoint, the attribute value appears as originally set.
Replica busy errors are not logged by default because they are usually benign. T o see the errors, turn
on the replication error logging, log level 8192. T he log levels are described in more detail in the
Directory Server Configuration and Command-Line Tool Reference.
9.6. Configuring Cascading Replication
Setting up cascading replication, as shown in Figure 9.4, “Cascading Replication”, has three major steps,
for each server in the scenario, the supplier on Server A, which holds a read-write replica; the
304
Chapter 9. Managing Replication
consumer/supplier on hub Server B, which holds a read-only replica; and the consumer on Server C,
which holds a read-only replica:
Section 9.6.1, “Configuring the Read-Write Replica on the Supplier Server”
Section 9.6.2, “Configuring the Read-Only Replica on the Consumer Server”
Section 9.6.3, “Configuring the Read-Only Replica on the Hub”
Section 9.6.4, “Setting up the Replication Agreements”
9.6.1. Configuring the Read-Write Replica on the Supplier Server
Next, configure the supplier server, which holds the original copy of the database:
1. Specify the supplier settings for the server.
a. In the Directory Server Console, select the Configuration tab.
b. In the navigation tree, select the Replication folder.
c. In the right-hand side of the window, select the Supplier Settings tab.
d. Check the Enable Changelog checkbox.
T his activates all of the fields in the pane below that were previously grayed out.
e. Specify a changelog by clicking the Use default button, or click the Browse button to
display a file selector.
f. Set the changelog parameters for the number and age of the log files.
Clear the unlimited checkboxes to specify different values.
g. Click Save.
2. Specify the replication settings required for a read-write replica.
a. In the navigation tree on the Configuration tab, expand the Replication node, and
highlight the database to replicate.
T he Replica Settings tab opens in the right-hand side of the window.
b. Check the Enable Replica checkbox.
c. In the Replica Role section, select the Single Master radio button.
305
Red Hat D irectory Server 8.2 Administration Guide
d. In the Com m on Settings section, specify a Replica ID, which is an integer between 1
and 65534 , inclusive.
T he replica ID must be unique for a given suffix, different from any other ID used for readwrite replicas on this server and on other servers.
e. In the Com m on Settings section, specify a purge delay in the Purge delay field.
T he purge delay is how often the state information stored for the replicated entries is
deleted.
f. Click Save.
After setting up the supplier replica, begin configuring the replication agreements.
9.6.2. Configuring the Read-Only Replica on the Consumer Server
1. Create the database for the read-only replica if it does not exist. See Section 2.1.1, “Creating
Suffixes” for instructions on creating suffixes.
2. Create the entry for the supplier bind DN on the consumer server if it does not exist. T he supplier
bind DN is the special entry that the supplier will use to bind to the consumer. T his is described in
Section 9.3, “Creating the Supplier Bind DN Entry”.
3. Specify the replication settings required for a read-only replica.
a. In the Directory Server Console, select the Configuration tab.
b. In the navigation tree, expand the Replication folder, and highlight the replica database.
T he Replica Settings tab for that database opens in the right-hand side of the window.
306
Chapter 9. Managing Replication
c. Check the Enable Replica checkbox.
d. In the Replica Role section, select the Dedicated Consum er radio button.
e. In the Com m on Settings section, specify a purge delay in the Purge delay field.
T his option indicates how often the state information stored for the replicated entries is
purged.
f. In the Update Settings section, specify the bind DN that the supplier will use to bind to
the replica. Enter the supplier bind DN in the Enter a new Supplier DN field, and click
Add. T he supplier bind DN appears in the Current Supplier DNs list.
T he supplier bind DN should be the entry created in step 2. T he supplier bind DN is a
privileged user because it is not subject to access control in the replicated database.
307
Red Hat D irectory Server 8.2 Administration Guide
NOTE
T here can be multiple supplier bind DNs per consumer but only one supplier DN per
replication agreement.
g. Specify the URL for any supplier servers to which to refer updates.
By default, all updates are first referred to the supplier servers that are specified here. If no
suppliers are set here, updates are referred to the supplier servers that have a replication
agreement that includes the current replica.
In cascading replication, referrals are automatically sent to the hub server, which in turn
refers the request to the original supplier. T herefore, set a referral to the original supplier to
replace the automatically generated referral.
4. Click Save.
Repeat these steps for every consumer server in the replication configuration, then configure the hub
replica.
9.6.3. Configuring the Read-Only Replica on the Hub
Do this to set up a hub, which receives replication updates from the supplier and propagates them to
consumers:
1. Create the database for the read-only replica if it does not exist. See Section 2.1.1, “Creating
Suffixes” for instructions on creating suffixes.
2. Create the entry for the supplier bind DN on the consumer server if it does not exist. T he supplier
bind DN is the special entry that the supplier will use to bind to the consumer. T his is described in
Section 9.3, “Creating the Supplier Bind DN Entry”.
3. Create the changelog for the hub server.
T he hub must maintain a changelog even though it does not accept update operations because it
records the changes sent from the supplier server.
a. In the Directory Server Console, select the Configuration tab.
b. In the navigation tree, select the Replication folder.
c. In the right-hand side of the window, select the Supplier Settings tab.
308
Chapter 9. Managing Replication
d. Check the Enable Changelog checkbox.
T his activates all of the fields in the pane below that were previously grayed out.
e. Specify a changelog by clicking the Use default button, or click the Browse button to
display a file selector.
f. Set the changelog parameters for the number and age of the log files.
Clear the unlimited checkboxes to specify different values.
g. Click Save.
4. Specify the required hub replica settings.
a. In the Directory Server Console, select the Configuration tab.
b. In the navigation tree, expand the Replication folder, and highlight the replica database.
T he Replica Settings tab for that database opens in the right-hand side of the window.
c. Check the Enable Replica checkbox.
309
Red Hat D irectory Server 8.2 Administration Guide
d. In the Replica Role section, select the Hub radio button.
e. In the Com m on Settings section, specify a purge delay in the Purge delay field.
T his option sets how often the state information stored for the replicated entries is purged.
f. In the Update Settings section, specify the bind DN that the supplier will use to bind to
the replica. Enter the supplier bind DN in the Enter a new Supplier DN field, and click
Add. T he supplier bind DN appears in the Current Supplier DNs list.
T he supplier bind DN should be the entry created in step 2. T he supplier bind DN is a
privileged user because it is not subject to access control in the replicated database.
NOTE
T here can be multiple supplier bind DNs per consumer but only one supplier DN per
replication agreement.
310
Chapter 9. Managing Replication
g. Specify the URL for any supplier servers to which to refer updates.
By default, all updates are first referred to the supplier servers that are specified here. If no
suppliers are set here, updates are referred to the supplier servers that have a replication
agreement that includes the current replica.
Automatic referrals assume that clients bind over a regular connection; this has a URL in
the form ldap://hostname:port. For clients to bind to the supplier using SSL, use this field
to specify a referral of the form ldaps://hostname:port, where the s in ldaps indicates a
secure connection.
5. Click Save.
When all the hubs are configured, then configure the supplier replica.
9.6.4. Setting up the Replication Agreements
Cascading replication requires two sets of replication agreements, the first between the supplier and the
hub and the second between the hub and the consumer. T o set up the replication agreements:
1. Create the replication agreement on the supplier for the hub, then use the supplier server to
initialize the replica on the hub server.
2. T hen create the replication agreement on the hub for each consumer, and initialize the consumer
replicas from the hub.
T o set up a replication agreement:
1. In the navigation tree of the Configuration tab, right-click the database to replicate, and select
New Replication Agreem ent.
Alternatively, highlight the database, and select New Replication Agreem ent from the
Object menu to start the Replication Agreement Wizard.
2. In the first screen, fill in a name and description for the replication agreement, and hit Next.
3. In the Source and Destination screen, fill in the URL for the consumer and the supplier bind
DN and password on that consumer. If the target server is not available, hit in other to fill in the
information manually.
311
Red Hat D irectory Server 8.2 Administration Guide
Unless there is more than one instance of Directory Server configured, by default, there are no
consumers available in the drop-down menu.
T he port listed is the non-SSL port, even if the Directory Server instance is configured to run
over SSL. T his port number is used only for identification of the Directory Server instance in
the Console; it does not specify the actual port number or protocol that is used for replication.
If SSL is enabled on the servers, it is possible to select the Using encrypted SSL
connection radio button for SSL client authentication. Otherwise, fill in the supplier bind DN
and password.
NOTE
If attribute encryption is enabled, a secure connection must be used for the encrypted
attributes to be replicated.
4. Select the connection type. T here are three options:
Use LDAP. T his sets a standard, unencrypted connection.
Use TLS/SSL. T his uses a secure connection over the server's secure LDAPS port, such as
636. T his setting is required to use T LS/SSL.
Use Start TLS. T his uses Start T LS to establish a secure connection over the server's
312
Chapter 9. Managing Replication
standard port.
NOTE
If secure binds are required for simple password authentication (Section 13.5.1, “Requiring
Secure Binds”), then any replication operations will fail unless they occur over a secure
connection. Using a secure connection (SSL/T LS and Start T LS connections or SASL
authentication) is recommended, anyway.
5. Select the appropriate authentication method and supply the required information. T his gives the
information that the supplier uses to authenticate and bind to the consumer server to send
updates.
Simple means that the server connects over the standard port with no encryption. T he only
required information is the bind DN and password for the Replication Manager (which must
exist on the consumer server).
Server TLS/SSL Certificate uses the supplier's SSL certificate to authenticate to the consumer
server. A certificate must be installed on the supplier for certificate-based authentication, and
the consumer server must have certificate mapping configured so that it can map the subject
DN in the supplier's certificate to its Replication Manager entry.
Configuring SSL and certificate mapping is described in Section 14.2, “Using T LS/SSL”.
SASL/DIGEST-MD5, like simple authentication, requires only the bind DN and password to
authenticate. T his can run over a standard or SSL/T LS connection.
SASL/GSSAPI requires the supplier server to have a Kerberos keytab (as in Section 14.3.4.2,
“About the KDC Server and Keytabs”), and the consumer server to have a SASL mapping to
map the supplier's principal to the real replication manager entry (as in Section 14.3.5.1,
“Configuring SASL Identity Mapping from the Console”).
6. Hit Next.
7. Fractional replication controls which entry attributes are replicated between servers. By default, all
attributes are replicated. T o select attributes that will not be replicated to the consumer, check the
Enable Fractional Replication checkbox. T hen, highlight the attribute (or attributes) in
the Included column on the right, and click Rem ove. All attributes that will not be replicated are
listed in the Excluded column on the left, as well as in the summary the replication agreement is
complete.
313
Red Hat D irectory Server 8.2 Administration Guide
8. Set the schedule for when replication runs. By default, replication runs continually.
314
Chapter 9. Managing Replication
NOTE
T he replication schedule cannot cross midnight (0000). So, it is possible to set a schedule
that begins at 0001 and ends at 2359 on the same day, but it is not possible to set one
that begins at 2359 on one day and ends at 0001 on the next.
Hit Next.
9. Set when the consumer is initialized. Initializing a consumer manually copies all data over from the
supplier to the consumer. T he default is to create an initialization file (an LDIF of all supplier data)
so that the consumer can be initialized later. It is also possible to initialize the consumer as soon
as the replication agreement is completed or not at all. For information on initializing consumers,
see Section 9.10, “Initializing Consumers”. For cascading replication, consider the following:
Create the supplier-hub replication agreement on the supplier first, and initialize the hub from
the supplier.
Create the hub-consumer replication agreements on the hub, and initialize the consumers from
the hub.
NOTE
Replication will not begin until the consumer is initialized.
IMPORTANT
For multi-master replication, be sure that consumers are only initialized once, by one
supplier. When checking the replication status, be sure to check the replication agreement
entry, on the appropriate supplier, which was used to initialize the consumer.
Hit Next.
10. T he final screen shows the settings for the replication agreement, as it will be included in the
315
Red Hat D irectory Server 8.2 Administration Guide
dse.ldif file. Hit Done to save the agreement.
NOTE
After creating a replication agreement, the connection type (SSL or non-SSL) cannot be change
because LDAP and LDAPS connections use different ports. T o change the connection type, recreate the replication agreement.
9.7. Configuring Replication from the Command Line
Replication can be configured on the command line by creating the appropriate replica and agreement
entries on the servers. T he process follows the same order as setting up replication through the
Directory Server Console:
1. Create the supplier bind DN on every consumer, hub, and multi-master supplier (Section 9.3,
“Creating the Supplier Bind DN Entry”).
2. If the corresponding database and suffix do not exist on one of the replicas, create it
(Section 2.1.1, “Creating Suffixes”).
3. Configure the supplier replicas (Section 9.7.1, “Configuring Suppliers from the Command Line”).
4. Configure consumers (Section 9.7.2, “Configuring Consumers from the Command Line”).
5. Configure hubs for cascading replication (Section 9.7.3, “Configuring Hubs from the Command
Line”).
6. Create the replication agreements (Section 9.7.4, “Configuring Replication Agreements from the
Command Line”). For cascading replication, create the agreement between the supplier and hub,
then between the hub and consumers; for multi-master, create the agreements between all
suppliers, then between the suppliers and consumers.
316
Chapter 9. Managing Replication
7. Lastly, initialize all of the consumers (Section 9.7.5, “Initializing Consumers Online from the
Command Line”), if the consumers were not initialized when the replication agreement was
created.
9.7.1. Configuring Suppliers from the Command Line
T here are two steps to setting up the supplier replica. First, the changelog must be enabled, which
allows the supplier to track changes to the Directory Server. T hen, the supplier replica is created.
1. On the supplier server, use ldapm odify to create the changelog entry.
Example 9.2. Example Changelog Entry
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389
-h supplier1.example.com -v -a
dn: cn=changelog5,cn=config
objectclass: top
objectclass: extensibleObject
cn: changelog5
nsslapd-changelogdir: /var/lib/dirsrv/slapd-instance_name/changelogdb
nsslapd-changelogmaxage: 10d
T here are two important attributes with the changelog.
nsslapd-changelogdir sets the directory where the changelog is kept.
nsslapd-changelogmaxage sets how long the changelog is kept; since the changelog can
get very large, this helps trim the changelog to prevent affecting server performance and using
up disk space. If this parameter is not set, the default is for the changelog to be kept forever.
T he changelog entry attributes are described in T able 9.1, “Changelog Attributes”. T hese
attributes are described in more detail in the Directory Server Configuration and Command-Line
Tool Reference.
2. Create the supplier replica.
Example 9.3. Example Supplier Replica Entry
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389
-h supplier1.example.com -v -a
dn: cn=replica,cn=dc=example\,dc=com,cn=mapping tree,cn=config
objectclass: top
objectclass: nsds5replica
objectclass: extensibleObject
cn: replica
nsds5replicaroot: dc=example,dc=com
nsds5replicaid: 7
nsds5replicatype: 3
nsds5flags: 1
nsds5ReplicaPurgeDelay: 604800
nsds5ReplicaBindDN: cn=replication manager,cn=config
nsds5replicaroot sets the subtree (suffix) which is being replicated.
nsds5replicatype sets what kind of replica this database is. For either a single master or a
multi-master supplier, this value must be 3.
nsds5replicaid sets the replica ID. T he value must be unique among all suppliers and hubs; the
317
Red Hat D irectory Server 8.2 Administration Guide
valid range is 1 to 65534 .
nsds5ReplicaPurgeDelay sets how long an entry holds its status information or how long a
tombstone entry is kept before deleting the information. T he default value is 604 800 (one
week).
nsds5flags sets whether the replica writes to the changelog. For a supplier, this value must be
1.
T he replica entry attributes are described in T able 9.2, “Replica Attributes”.
After creating every supplier which will take part in the replication setup, then begin creating the
replication agreements.
T able 9.1. Changelog Attributes
Object Class or Attribute
Description
objectclass: top
Required object class for every
entry.
objectclass: extensibleObject
An object class which allows
any other object class or
attribute to be added to an entry.
cn: changelog5
T he naming attribute for the
changelog entry.
nsslapd-changelogdir: directory
Sets the file and directory
Any directory; the default is
changelog, to which the
/var/lib/dirsrv/slapd-in
Directory Server writes changes. stance_name/changelogdb.
nsslapd-changelogmaxage:
number unit
Sets how long the changelog
keeps an entry before purging it.
T his is not used by consumers,
but is recommended for hubs
and suppliers, which keep
changelogs.
318
Values
Any string; the default usage is
to set the common name to
changelog5.
T he number is any integer.
T he unit can be s for seconds,
m for minutes, h for hours, d for
days, or w for weeks.
Chapter 9. Managing Replication
T able 9.2. Replica Attributes
Object Class or Attribute
Description
Values
objectclass: top
Required object class for every
entry.
objectclass: extensibleObject
An object class which allows
any other object class or
attribute to be added to an entry.
objectclass: nsds5replica
An object class which allows
replication attributes to be
added to an entry.
cn: replica
T he naming attribute for the
replica.
Any string; the default usage is
to set the common name to
replica for every configured
replica.
nsds5replicaroot: suffix
Sets which subtree is replicated.
A root suffix associated with a
database, since the entire
database is replicated. For
example:
dc=example,dc=com
nsds5replicaid: number
nsds5replicatype: number
T he ID of the replica. T his must
be set to 65535 for consumers
or hubs. For suppliers, this
value must be a unique value.
Sets the type of replica, either
read-only or read-write.
1 to 65534 , inclusive, for
suppliers.
65535 for consumers and
hubs.
2 for consumers and hubs
(read-only replicas)
3 for both single and multimaster suppliers (read-write
replicas)
nsds5flags: number
Sets whether the replica writes
to the changelog.
0 means the replica does not
write to the changelog; this is
the default for consumers.
1 means the relics writes to the
changelog; this is the default for
hubs and suppliers.
nsds5ReplicaPurgeDelay:
number
Sets the period of time in
seconds to wait before purging
the state information from an
entry or purging tombstone
entries. T his setting is required
for all types of replicas —
suppliers, hubs, and consumers.
0 (keep forever) to
214 74 8364 7 (the maximum
32-bit integer); the default value
is 604 800, one week.
nsds5ReplicaBindDN: DN
T he supplier bind DN used by
Any DN; the recommended DN
319
Red Hat D irectory Server 8.2 Administration Guide
the supplier to bind to the
consumer. T his is required for
consumers, hubs, and multimaster suppliers, but not for
single-master suppliers.
is cn=Replication
Manager,cn=config.
NOTE
For security, it is strongly
recommended that you
do not use the Directory
Manager as the supplier
bind DN.
nsds5replicareferral: URL
Optional. An LDAP URL which a
consumer or hub to which a
consumer or hub can forward
update requests. By default,
update requests are sent to the
masters for the consumer; use
this parameter to override the
default.
Any LDAP URL. For example:
nsds5replicareferral:
ldap://supplier1.exampl
e.com:389
.
9.7.2. Configuring Consumers from the Command Line
On the consumer host, such as consum er1.exam ple.com , create the replica entry. T his entry
identifies the database and suffix as participating in replication and sets what kind of replica the
database is. T here are four key attributes:
nsds5replicaroot sets the subtree (suffix) which is being replicated.
nsds5replicatype sets what kind of replica this database is. For a consumer, this value must be 2.
nsds5ReplicaBindDN give the DN as which the supplier will bind to the consumer to make changes.
nsds5flags sets whether the replica writes to the changelog. For a consumer, this value must be 0.
T his ldapm odify creates a new consumer replica on the consum er1.exam ple.com host for the
dc=exam ple,dc=com subtree.
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
consumer1.example.com -v -a
dn: cn=replica,cn=dc=example\,dc=com,cn=mapping tree,cn=config
objectclass: top
objectclass: nsds5replica
objectclass: extensibleObject
cn: replica
nsds5replicaroot: dc=example,dc=com
nsds5replicatype: 2
nsds5ReplicaBindDN: cn=replication manager,cn=config
nsds5flags: 0
T he replica entry attributes are described in T able 9.2, “Replica Attributes”. T hese attributes are
described in more detail in the Directory Server Configuration and Command-Line Tool Reference.
9.7.3. Configuring Hubs from the Command Line
Hubs are intermediate read-only replicas which receive updates from suppliers and pass them on to
320
Chapter 9. Managing Replication
other consumers. T hese are part of the cascading replication scenario, described in Section 9.2.3,
“Cascading Replication”.Creating the hub has two steps: first, creating the changelog database since the
hub keeps a record of changes sent by the supplier, and second, configuring the hub replica.
1. On the hub server, such as hub1.exam ple.com , use ldapm odify to create the changelog
entry.
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
hub1.example.com -v -a
dn: cn=changelog5,cn=config
objectclass: top
objectclass: extensibleObject
cn: changelog5
nsslapd-changelogdir: /var/lib/dirsrv/slapd-instance_name/changelogdb
T here is one important attribute with the changelog, nsslapd-changelogdir, which sets the
directory where the changelog is kept.
T he changelog entry attributes are described in T able 9.1, “Changelog Attributes”. T hese
attributes are described in more detail in the Directory Server Configuration and Command-Line
Tool Reference.
2. On the hub host, create the replica entry. T his ldapm odify command creates a new hub replica
on the hub1.exam ple.com host for the dc=exam ple,dc=com subtree.
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
hub1.example.com -v -a
dn: cn=replica,cn=dc=example\,dc=com,cn=mapping tree,cn=config
objectclass: top
objectclass: nsds5replica
objectclass: extensibleObject
cn: replica
nsds5replicaroot: dc=example,dc=com
nsds5replicatype: 2
nsds5ReplicaPurgeDelay: 604800
nsds5ReplicaBindDN: cn=replication manager,cn=config
nsds5flags: 1
T his entry identifies the database and suffix as participating in replication and sets what kind of
replica the database is. T here are five key attributes:
nsds5replicaroot sets the subtree (suffix) which is being replicated.
nsds5replicatype sets what kind of replica this database is. For a hub, this value must be 2.
nsds5ReplicaPurgeDelay sets how long an entry holds its status information or how long a
tombstone entry is kept before deleting the information. T he default value is 604 800 (one
week).
nsds5ReplicaBindDN give the DN as which the supplier will bind to the hub to make changes.
nsds5flags sets whether the replica writes to the changelog. For a hub, this value must be 1.
T he replica entry attributes are described in T able 9.2, “Replica Attributes”. T hese attributes are
described in more detail in the Directory Server Configuration and Command-Line Tool Reference.
9.7.4. Configuring Replication Agreements from the Command Line
When setting up replication agreements, first set them up between all suppliers, then between the
suppliers and the hubs, and last between the hub and the consumers.
321
Red Hat D irectory Server 8.2 Administration Guide
T he replication agreement has to define eight separate attributes:
T he consumer host (nsds5replicahost) and port (nsds5replicaport).
T he DN for the supplier to use to bind with the consumer (nsds5ReplicaBindDN).
T he way that the supplier binds (nsds5replicabindmethod).
Any credentials required (nsds5replicabindcredentials) for that bind method and specified DN.
T he subtree being replicated (nsds5replicaroot).
T he replication schedule (nsds5replicaupdateschedule).
Any attributes which will not be replicated (nsds5replicatedattributelist).
Use ldapm odify to add a replication agreement to every supplier for every consumer which it will
updated. For example:
Example 9.4 . Example Replication Agreement Entry
dn: cn=ExampleAgreement,cn=replica,cn=dc=example\,dc=com,cn=mapping
tree,cn=config
objectclass: top
objectclass: nsds5replicationagreement
cn: ExampleAgreement
nsds5replicahost: consumer1
nsds5replicaport: 389
nsds5ReplicaBindDN: cn=replication manager,cn=config
nsds5replicabindmethod: SIMPLE
nsds5replicaroot: dc=example,dc=com
description: agreement between supplier1 and consumer1
nsds5replicaupdateschedule: 0000-0500 1
nsds5replicatedattributelist: (objectclass=*) $ EXCLUDE authorityRevocationList
nsds5replicacredentials: secret
nsds5BeginReplicaRefresh: start
T he replication agreement attributes are listed in T able 9.3, “Replication Agreement Attributes”. T hese
attributes are described in more detail in the Directory Server Configuration and Command-Line Tool
Reference.
After creating every replication agreement, begin initializing consumers.
322
Chapter 9. Managing Replication
T able 9.3. Replication Agreement Attributes
Object Class or Attribute
Description
objectclass: top
Required object class for every
entry.
objectclass:
nsds5replicationagreement
An operational object class
which contains the replication
agreement attributes.
cn: agreement_name
T he naming attribute for the
replication agreement.
Any string.
nsds5replicahost: hostname
Gives the hostname of the
consumer server; the hostname
can be the fully qualified host
and domain name. If T LS/SSL is
enabled, the fully-qualified
domain name is required.
Any hostname. For example:
Gives the LDAP port for the
consumer server.
Any port number.
nsds5replicaport: number
Values
nsds5replicahost:
consumer1
T o use T LS/SSL, give the
secure port number (636 by
default) and set the
nsds5ReplicaTransportInfo
attribute to SSL.
T o use Start T LS, which initiates
a secure connection over a
standard port, use the standard
port, 389, with the
nsds5ReplicaTransportInfo
attribute to T LS.
T o use simple authentication,
use the standard port, 389, with
the
nsds5ReplicaTransportInfo
attribute to LDAP.
nsds5replicatransportinfo:
method
T o use T LS/SSL, set this
parameter to SSL.
T o use Start T LS, which initiates
a secure connection over a
standard port, set this
parameter to T LS.
T o use simple authentication,
set this parameter to LDAP.
nsds5ReplicaBindDN: DN
T he supplier bind DN used by
the supplier to bind to the
consumer. T his is required for
consumers, hubs, and multimaster suppliers, but not for
empty or LDAP (standard
connection over port 389)
SSL (secure connection over the
secure port, such as 636)
T LS (secure connection over
the standard port, 389, using
Start T LS)
Any DN; the recommended DN
is cn=Replication
Manager,cn=config.
323
Red Hat D irectory Server 8.2 Administration Guide
single-master suppliers.
nsds5replicabindmethod: type
T he connection type for
replication between the servers.
T he connection type defines
how the supplier authenticates
to the consumer.
Leaving the bind method empty
or setting it to SIMPLE means
that the server uses basic
password-based authentication.
T his requires the
nsds5ReplicaBindDN and
nsds5ReplicaCredentials
attributes to give the bind
information.
T he SSLCLIENT AUT H option
uses a secure connection. T his
requires setting the
nsds5ReplicaTransportInfo
attribute be set to SSL or T LS.
For certificate-based
authentication, the consumer
server must also have a
certificate mapping to map the
subject DN in the supplier's
certificate to the replication
manager entry.
Using SASL/GSSAPI requires
that the supplier server have a
Kerberos keytab (as in
Section 14.3.4.2, “About the
KDC Server and Keytabs”), and
the consumer server have a
SASL mapping to map the
supplier's principal to the real
replication manager entry (as in
Section 14.3.5.1, “Configuring
SASL Identity Mapping from the
Console”).
T he SASL/DIGEST -MD5
setting, like SIMPLE, uses
password-based authentication
and requires the
nsds5ReplicaBindDN and
nsds5ReplicaCredentials
attributes to give the bind
information.
324
SIMPLE
SSLCLIENT AUT H
SASL/GSSAPI
SASL/DIGEST -MD5
Chapter 9. Managing Replication
NOTE
If secure binds are
required for simple
password authentication
(Section 13.5.1,
“Requiring Secure
Binds”), then any
replication operations will
fail unless they occur
over a secure
connection. Using a
secure connection
(SSL/T LS and Start T LS
connections or SASL
authentication) is
recommended, anyway.
nsds5replicabindcredentials:
hash
Only for simple authentication.
Stores the hashed password
used with the bind DN given for
simple authentication.
nsds5replicaroot: suffix
Sets which subtree is replicated.
A root suffix associated with a
database, since the entire
database is replicated. For
example:
dc=example,dc=com
description: text
A text description of the
replication agreement.
nsds5replicatedattributelist:
'(objectclass=*)' $ EXCLUDE
attributes
Optional. Sets which attributes
will not be replicated. T he filter
must be set to
"(objectclass=* )", and the
list of attributes are separated
by a single space.
nsds5replicaupdateschedule:
start_time end_time days
Sets the start and end time for
the replication updates and the
days on which replication
occurs. If the schedule is
omitted, replication will take
place all the time.
T he replication schedule cannot
cross midnight (0000). So, it is
possible to set a schedule that
begins at 0001 and ends at
2359 on the same day, but it is
Any text string. It is advisable to
make this a useful description,
such as agreement between
supplier1 and consumer1.
'(objectclass=*)' $
EXCLUDE userPassword
manager cn
Has the following value, with the
start (SSSS) and end (EEEE)
times set in the form HHMM
T he times are given in 24 hour
clock format, so 0000 is
midnight and 2359 is 11:59 PM.
For example, the setting 1030
1630 schedules replication from
10:30 AM to 4:30 PM. T he times
cannot wrap around midnight, so
the setting 2300 0100 is not
325
Red Hat D irectory Server 8.2 Administration Guide
not possible to set one that
begins at 2359 on one day and
ends at 0001 on the next.
valid.
T he days ranging from 0
(Sunday) to 6 (Saturday).
Setting 06 schedules replication
on Sunday and Saturday, while
135 schedules replication on
Monday, Wednesday, and
Friday.
nsds5replicaupdatesched
ule:
SSSS EEEE DDDDDDD
For example, this schedules
replication between midnight
(0000) and 5am (0500) on
Monday and T uesday:
nsds5replicaupdatesched
ule:
0000 0500 12
nsds5BeginReplicaRefresh:
start
Optional. Performs an online
(immediate) initialization of the
consumer. If this is set, the
attribute is only present as long
as the consumer is being
initialized; when the initialization
is complete, the attribute is
deleted automatically.
start
T o initialize the consumer, this
attribute must have a value of
start; any other value is
ignored.
If this is not set, then consumer
initialization must be performed
manually.
9.7.5. Initializing Consumers Online from the Command Line
An online initialization can be initiated from the command line by adding the nsds5replicarefresh
attribute to the replication agreement entry. If the attribute is included when the replication agreement is
created, initialization begins immediately. It can be added later to initialize the consumer at any time. T his
attribute is absent by default, and it will be automatically deleted once the consumer initialization is
complete.
1. Find the DN of the replication agreement on the supplier server that is for the consumer to be
initialized. For example:
ldapsearch -h supplier1.example.com -p 389 -D "cn=directory manager" -w
secret -s sub -b cn=config "(objectclass=nsds5ReplicationAgreement)"
T his command returns all of the replication agreements configured on the supplier in LDIF format.
Get the DN of the replication agreement with the consumer to be initialized. T his is the replication
agreement which will be edited.
326
Chapter 9. Managing Replication
2. Edit the replication agreement, and add the nsds5BeginReplicaRefresh attribute:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
supplier1.example.com
dn: cn=ExampleAgreement,cn=replica,cn=dc=example\,dc=com,cn=mapping
tree,cn=config
changetype: modify
replace: nsds5beginreplicarefresh
nsds5beginreplicarefresh: start
ldapm odify does not prompt for input; simply type in the LDIF statement, and then hit enter
twice when the LDIF statement is complete. Close the ldapm odify utility by hitting Ctrl+C.
When the initialization is complete, the nsds5beginreplicarefresh attribute is automatically deleted
from the replication agreement entry.
IMPORTANT
For multi-master replication, be sure that consumers are only initialized once, by one supplier.
When checking the replication status, be sure to check the replication agreement entry, on the
appropriate supplier, which was used to initialize the consumer.
Initializing consumers from the command line is also explained in Section 9.10.3, “Initializing Consumers
Online Using the Command Line”. Manually initializing consumers is explained in Section 9.10.4, “Manual
Consumer Initialization Using the Command Line”. T he replication monitoring attributes are described in
more detail in the Directory Server Configuration and Command-Line Tool Reference.
NOTE
For large databases, the nsslapd-idletimeout setting must be set to a large enough time
period (or even an unlimited period) to allow the entire database to be initialized before the
operation times out. Alternatively, the nsIdleTimeout setting for the supplier bind DN entry can
be set high enough to allow the online initialization operation to complete, without having to
change the global setting.
T o keep data integrity, initialize the consumer databases from the appropriate supplier. Determining the
correct supplier can be more difficult in mixed replication environments, but, even when manually
initializing consumers, consider four things:
Use one supplier, a data master, as the source for initializing consumers.
Do not reinitialize a data master when the replication agreements are created. For example, do not
initialize server1 from server2 if server2 has already been initialized from server1.
For a multi-master scenario, initialize all of the other master servers in the configuration from one
master.
For cascading replication, initialize all of the hubs from a supplier, then initialize the consumers from
the hubs.
9.8. Making a Replica Updatable
327
Red Hat D irectory Server 8.2 Administration Guide
Making a read-only server writable means changing the replica from a dedicated consumer or a hub to a
supplier.
1. Make sure there are no updates in progress.
2. Stop the supplier server.
3. Open the Directory Server Console for the read-only replica.
4. In the Configuration tab, select Replication. In the right pane, select the Enable
changelog checkbox.
5. Select the suffix, and, in the Replica Settings tab, change the replica role to a single master
or multi-master.
6. Assign a unique replica ID.
7. Save the changes, and restart the server.
9.9. Deleting the Changelog
T he changelog is a record of all modifications on a given replica that the supplier uses to replay these
modifications to replicas on consumer servers (or suppliers in the case of multi-master replication).
If a supplier server goes offline, it is important to be able to delete the changelog because it no longer
holds a true record of all modifications and, as a result, should not be used as a basis for replication. A
changelog can be effectively deleted by deleting the log file or by moving it to a new location.
After deleting the changelog, reinitialize the consumers and begin the replication process afresh.
Section 9.9.1, “Removing the Changelog”
Section 9.9.2, “Moving the Changelog to a New Location”
9.9.1. Removing the Changelog
T o remove the changelog from the supplier server:
1. In the Directory Server Console, select the Configuration tab.
328
Chapter 9. Managing Replication
2. Select the Replication folder in the left navigation tree and then the Supplier Server
Settings tab in the right pane.
3. Clear the Enable Changelog checkbox.
4. Click Save.
5. Restart the Directory Server.
6. Reinitialize the consumers, as in Section 9.10, “Initializing Consumers”.
NOTE
If the changelog is removed, the consumer servers must be reinitialized.
9.9.2. Moving the Changelog to a New Location
Sometimes, it may be necessary to delete the changelog while the supplier server is still running and
continuing to log changes.
In this case, simply move the current changelog file to a new location. T he server then automatically
creates a new changelog in the specified directory.
After moving the changing, reinitialize the consumers; any changes to the changelog requires consumer
reinitialization.
9.10. Initializing Consumers
Once a replication agreement is created, the consumer must be initialized; that is, the data must be
physically copied from the supplier server to the consumer servers.
NOTE
Replication will not begin until the consumer is initialized.
Section 9.10.1, “When to Initialize a Consumer”
Section 9.10.2, “Online Consumer Initialization Using the Console”
Section 9.10.3, “Initializing Consumers Online Using the Command Line”
Section 9.10.4, “Manual Consumer Initialization Using the Command Line”
Section 9.10.5, “Filesystem Replica Initialization”
329
Red Hat D irectory Server 8.2 Administration Guide
NOTE
For large databases, the nsslapd-idletimeout setting must be set to a large enough time
period (or even an unlimited period) to allow the entire database to be initialized before the
operation times out. Alternatively, the nsIdleTimeout setting for the supplier bind DN entry can
be set high enough to allow the online initialization operation to complete, without having to
change the global setting.
9.10.1. When to Initialize a Consumer
Consumer initialization involves copying data from the supplier server to the consumer server. Once the
subtree has been physically placed on the consumer, the supplier server can begin replaying update
operations to the consumer server.
Under normal operations, the consumer should not ever have to be reinitialized. However, any time there
is a chance that there is a big discrepancy between the supplier's data and the consumer's, reinitialize
the consumer. For example, if the data on the supplier server is restored from backup, then all
consumers supplied by that server should be reinitialize. As another example, if the supplier has not
been able to contact the consumer for a long time, like a week, the supplier may determine that the
consumer is too far out of date to be updated, and must be reinitialized.
T he consumer can either be initialized online using the Console or manually using the command line.
Online consumer initialization using the Console is an effective method of initializing a small number of
consumers. However, since each replica is initialized in sequence, this method is not suited to initializing
a large number of replicas. Online consumer initialization is the method to use when the consumer is
initialized as part of configuring the replication agreement on the supplier server.
Manual consumer initialization using the command line is a more effective method of initializing a large
number of consumers from a single LDIF file.
9.10.2. Online Consumer Initialization Using the Console
Online consumer initialization using the Console is the easiest way to initialize or reinitialize a consumer.
However, for replicating across a slow link, this process can be very time-consuming, and manual
consumer initialization using the command line may be a more efficient approach. T his is described in
more detail Section 9.10.4, “Manual Consumer Initialization Using the Command Line”.
NOTE
When a consumer server is being initialized using the online consumer creation method, all
operations (including searches) on the replica are referred to the supplier server until the
initialization process is completed.
T o initialize or reinitialize a consumer online:
1. Create a replication agreement.
2. On the supplier server, on the Directory Server Console, select the Configuration tab.
3. Expand the Replication folder, then expand the replicated database. Right-click the replication
agreement, and choose Initialize Consum er from the pop-up menu.
330
Chapter 9. Managing Replication
A message opens warning that any information already stored in the replica on the consumer will
be removed.
4. Click Yes in the confirmation box.
Online consumer initialization begins immediately. T o check the status of the online consumer
initialization, open the Sum m ary tab in the Status box. If online consumer initialization is in progress,
the status shows that a replica is being initialized.
IMPORTANT
For multi-master replication, be sure that consumers are only initialized once, by one supplier.
When checking the replication status, be sure to check the replication agreement entry, on the
appropriate supplier, which was used to initialize the consumer.
T o update this window, right-click the replicated database icon in the navigation tree, and choose
Refresh Replication Agreem ents. When online consumer initialization finishes, the status
changes to reflect this.
For more information about monitoring replication and initialization status, see Section 9.17, “Monitoring
Replication Status”.
9.10.3. Initializing Consumers Online Using the Command Line
Online consumer initialization can be performed through the command line by adding the
nsds5BeginReplicaRefresh attribute to the replication agreement entry. T his attribute is absent by
default, and it will be automatically deleted once the consumer initialization is complete.
1. Find the DN of the replication agreement on the supplier server that is for the consumer to be
initialized. For example:
ldapsearch -h supplier1.example.com -p 389 -D "cn=directory manager" -w
secret -s sub
-b cn=config "(objectclass=nsds5ReplicationAgreement)"
T his command returns all of the replication agreements configured on the supplier in LDIF format.
Get the DN of the replication agreement with the consumer to be initialized. T his is the replication
agreement which will be edited.
331
Red Hat D irectory Server 8.2 Administration Guide
2. Edit the replication agreement, and add the nsds5BeginReplicaRefresh attribute:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
supplier1.example.com
dn: cn=ExampleAgreement,cn=replica,cn=dc=example\,dc=com,cn=mapping
tree,cn=config
changetype: modify
replace: nsds5beginreplicarefresh
nsds5beginreplicarefresh: start
ldapm odify does not prompt for input; simply type in the LDIF statement, and then hit enter
twice when the LDIF statement is complete. Close the ldapm odify utility by hitting Ctrl+C.
T o check the initialization status, do an ldapsearch for the replication agreement entry.
ldapsearch -D "cn=directory manager" -w secret -p 389 -h server.example.com -s
base -b 'cn=ExampleAgreement,cn=dc=example\,dc=com,cn=mapping tree,cn=config'
'(objectclass=*)'
If the nsds5BeginReplicaRefresh attribute is present, the initialization is still in progress. If the
initialization is complete, then the attribute nsds5ReplicaLastInitStatus shows the status. If the
initialization was successful, the value of nsds5ReplicaLastInitStatus is T otal update
succeeded. If the initialization was not successful, this attribute shows information about the error;
check the error logs for both the supplier and consumer for additional information.
IMPORTANT
For multi-master replication, be sure that consumers are only initialized once, by one supplier.
When checking the replication status, be sure to check the replication agreement entry, on the
appropriate supplier, which was used to initialize the consumer.
T he replication monitoring attributes are described in more detail in the Directory Server Configuration
and Command-Line Tool Reference.
9.10.4. Manual Consumer Initialization Using the Command Line
Manual consumer initialization using the command line is the fastest method of consumer initialization
for sites that are replicating very large numbers of entries. However, the manual consumer initialization
process is more complex than the online consumer initialization process. Red Hat suggests using the
manual process whenever the online process is inappropriate due to performance concerns.
Initializing or reinitializing a server manually has three steps:
1. Create a replication agreement.
2. Export the replica on the supplier server to an LDIF file.
See Section 9.10.4.1, “Exporting a Replica to LDIF”.
3. Import the LDIF file with the supplier replica contents to the consumer server.
See Section 9.10.4.2, “Importing the LDIF File to the Consumer Server”.
9.10.4 .1. Exporting a Replica to LDIF
T here are three ways to convert a replica database to LDIF:
332
Chapter 9. Managing Replication
When creating a replication agreement, by selecting Create consum er initialization file
in the Initialize Consum er dialog box of the Replication Agreement Wizard.
From the Directory Server Console, by right-clicking the replication agreement under the
Replication folder and choosing Create LDIF File from the pop-up menu.
From the command line by using the export command, as described in Section 4.2.3, “Exporting to
LDIF from the Command Line”. Exporting to LDIF with any of the command-line tools requires using an
option to export the database as a replica; this means that the exported LDIF contains the proper
entries to initialize the consumer when the LDIF is imported.
For the db2ldif and db2ldif.pl scripts, this is the -r option. For example:
db2ldif -r -n database1 -a /export/output.ldif
For the cn=export,cn=tasks,cn=config entry, this is the nsExportReplica attribute.
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389 h server.example.com
dn: cn=example export,cn=export,cn=tasks,cn=config
objectclass: extensibleObject
cn: example export
nsInstance: userRoot
nsFilename: /home/files/example.ldif
nsExportReplica: true
9.10.4 .2. Importing the LDIF File to the Consumer Server
Import the LDIF file which contains the supplier replica contents to the consumer server by using the
import features in the Directory Server Console or by using either the ldif2db script or ldif2db.pl
script. Both import methods are described in Section 4.1.4, “Importing from the Command Line”.
NOTE
With the ldif2db.pl script, the LDIF file import operation does not require a server restart. For
more information on command-line scripts, see the Directory Server Configuration and CommandLine Tool Reference.
9.10.5. Filesystem Replica Initialization
A very large database, such as one with several million entries, can take an hour or more to initialize a
consumer from the Console or even with manual initialization. T o save time, use filesystem replica
initialization.
Directory Server has the capability to initialize a replica using the database files from the supplier server.
T his avoids the need to rebuild the consumer database and can be done at essentially the speed of the
network between the two servers by transferring the files with FT P or NFS, for example. Instead of
sending entries via LDAP to replica servers, filesystem replica initialization populates the new database
on the destination server by backing up the supplier database on one server and restoring the database
on the destination server.
T his method of initializing consumers is especially useful in replication over wide-area networks or over
networks with slow or unstable connections.
For smaller databases, Red Hat recommends using manual initialization or initialize consumers from the
333
Red Hat D irectory Server 8.2 Administration Guide
Console.
NOTE
T he destination server must have the same architecture and the same bit size as the supplier
server for the initialization to succeed. For example, Red Hat Enterprise Linux 5 (64-bit) to Red
Hat Enterprise Linux 5 (64-bit).
9.10.5.1. Initializing the Consumer Replica from the Backup Files
1. Create a new database on the destination server to match the database from the source server.
Before initializing the consumer from the backup files, be certain that the appropriate database
has been created on the destination server so that the database exists to be restored and
initialized.
2. Enable replication on the backend as a dedicated consumer.
3. If there is already a replication agreement to that host and port, then replication should resume
immediately after running the restore script. Otherwise, create the replication agreement on the
source server (or whatever server is the supplier), and select the Do not initialize
consum ers at this tim e radio button.
4. Stop the source Directory Server if it is running. For example:
service dirsrv stop slapd-example
5. From the command line, run the db2bak utility, and archive the current directory installation.
/etc/dirsrv/slapd-instance_name/db2bak
In addition, a new backup can be created by hitting the Back Up Directory Server button in
the T asks tab of the Directory Server Console and then putting the name of the archive directory
in the Directory:... field. Alternatively, select any previous back-up to initialize the consumer.
T his backup information is recovered in the destination replica.
6. Restart the source Directory Server. For example:
service dirsrv start slapd-example
7. Copy the archived files to a directory on the destination server, such as archiveDirectory.
8. Stop the destination Directory Server if it is running.
service dirsrv stop slapd-example2
9. On the destination server, restore the archives with the bak2db script, using the optional -n
parameter to specify the backend instance name. T his -n parameter is similar to the -n used with
ldif2db and db2ldif. For example:
/usr/lib/dirsrv/slapd-example2/bak2db /tmp/archiveDirectory -n userRoot
10. Restart the destination Directory Server. For example:
service dirsrv start slapd-example2
334
Chapter 9. Managing Replication
Replication will begin on schedule as soon as the destination server is restarted.
For more information on using these scripts, see the Directory Server Configuration and Command-Line
Tool Reference.
9.11. Forcing Replication Updates
When a Directory Server involved in replication is stopped for regular maintenance, it must be updated
immediately when it comes back online. In the case of a supplier in a multi-master environment, the
directory information needs to be updated by the other supplier in the multi-master set. In other cases,
when a hub or a dedicated consumer is taken offline for maintenance, when they come back online, they
need to be updated by the supplier server.
Even if the replication agreements are configured to keep the supplier and consumer servers always in
sync, it is not sufficient to bring back up-to-date a server that has been offline for over five minutes. T he
Always Keep in Sync option means that the server generates a replication operation for every
update operation it processes. However, if this replication operation cannot be performed because the
consumer is offline, the operation times out after 10 minutes.
NOTE
T he procedures described in this section can only be used when replication is already set up and
consumers have been initialized.
T o ensure that directory information will be synchronized immediately when a server comes back online,
use either the Directory Server Console on the supplier server that holds the reference copy of the
directory information or a customizable script.
9.11.1. Forcing Replication Updates from the Console
T o ensure that replication updates are sent immediately when a consumer or a supplier in a multimaster replication configuration comes back online after a period of time, do the following on the supplier
server that holds the most recent version of the directory information:
1. In the Directory Server Console, click the Configuration tab, expand the Replication folder
and database nodes, and select the replication agreement corresponding to the replica to update.
2. Right click the replication agreement, and choose Send Updates Now from the drop-down list.
335
Red Hat D irectory Server 8.2 Administration Guide
T his initiates replication toward the server that holds the information that needs to be updated.
9.11.2. Forcing Replication Updates from the Command Line
From the consumer that requires updating, run a script that prompts the supplier to send replication
updates immediately. T his script is shown in Example 9.5, “Replicate_Now Script Example”.
Copy this example script and name it something like replicate_now.sh. Substitute the actual values
for the variables listed in Example 9.5, “Replicate_Now Script Example”.
NOTE
T his script must be run manually since it cannot be configured to run automatically as soon as
the server, which was offline, comes back online again.
336
Chapter 9. Managing Replication
Example 9.5. Replicate_Now Script Example
#!/bin/sh
SUP_HOST=supplier_hostname
SUP_PORT=supplier_portnumber
SUP_MGRDN=supplier_directoryManager
SUP_MGRPW=supplier_directoryManager_password
MY_HOST=consumer_hostname
MY_PORT=consumer_portnumber
ldapsearch -1 -T -h ${SUP_HOST} -p ${SUP_PORT} -D "${SUP_MGRDN}" \
-w ${SUP_MGRPW} -b "cn=mapping tree,cn=config"
\"(&(objectclass=nsds5replicationagreement)(nsDS5ReplicaHost=${MY_HOST})
\(nsDS5ReplicaPort=${MY_PORT}))" dn nsds5ReplicaUpdateSchedule > /tmp/$$
cat /tmp/$$ |awk 'BEGIN { s = 0 }/^dn: / { print $0;print "changetype:
modify";print
"replace: nsds5ReplicaUpdateSchedule";print "nsds5ReplicaUpdateSchedule:
0000-2359
0123456";print "-";print "";print $0;print "changetype: modify";
print "replace:nsds5ReplicaUpdateSchedule";}
/^nsds5ReplicaUpdateSchedule: / { s = 1; print $0; }/^$/{if ( $s == 1 ){ print
"-" ;
print ""; }else{ print "nsds5ReplicaUpdateSchedule: 0000-2359
0123456";print "-" ;
print ""; };s = 0; }
' > /tmp/ldif.$$echo "Ldif is in /tmp/ldif.$$"echo
ldapmodify -c -h ${SUP_HOST} -p ${SUP_PORT} -D "${SUP_MGRDN}" \-w ${SUP_MGRPW}
-f /tmp/ldif.$$
T able 9.4 . Replicate_Now Variables
Variable
Definition
supplier_hostname
Hostname of the supplier to contact for
information on replication agreements with the
current consumer.
supplier_portnumber
LDAP port in use on the supplier.
supplier_directoryManager
DN of the privileged Directory Manager user on
the supplier.
supplier_directoryManager_password
Password of the privileged Directory Manager
user on the supplier.
consumer_hostname
Hostname of the current consumer.
consumer_portnumber
LDAP port in use on the consumer.
For the update operation to occur over an SSL connection, modify the ldapm odify command in the
script with the appropriate parameters and values. For more information on the ldapm odify command,
see Section 3.2, “Managing Entries from the Command Line”.
9.12. Replication over SSL
337
Red Hat D irectory Server 8.2 Administration Guide
T he Directory Servers involved in replication can be configured so that all replication operations occur
over an SSL connection. T o use replication over SSL:
Configure both the supplier and consumer servers to use SSL.
Configure the consumer server to recognize the supplier server's certificate as the supplier DN. Do
this only to use SSL client authentication rather than simple authentication.
T hese procedures are described in Section 14.2, “Using T LS/SSL”.
If attribute encryption is enabled, a secure connection is required for replication.
NOTE
Replication configured over SSL with certificate-based authentication will fail if the supplier's
certificate is only capable of behaving as a server certificate, and not also a client during an SSL
handshake. Replication with certificate-based authentication uses the Directory Server's server
certificate for authentication to the remote server.
When the servers are configured to use SSL, configure an SSL connection for replication in the
Replication Agreement Wizard. T he Source and Destination sets how to bind between the
supplier and the consumer, and this is where SSL is set.
T here are two ways to use SSL for replication:
Select SSL Client Authentication.
With SSL client authentication, the supplier and consumer servers use certificates to authenticate to
each other.
Select Sim ple Authentication.
With simple authentication, the supplier and consumer servers use a bind DN and password to
authenticate to each other, which are supplied in the Replication Agreement Wizard text fields
provided. Simple authentication takes place over a secure channel but without certificates.
NOTE
If secure binds are required for simple password authentication (Section 13.5.1, “Requiring
Secure Binds”), then any replication operations will fail unless they occur over a secure
connection. Using a secure connection (SSL/T LS and Start T LS connections or SASL
authentication) is recommended, anyway.
Once a replication agreement is created, the connection type (SSL or non SSL) cannot be changed in the
agreement because LDAP and LDAPS connections use different ports. T o change the connection type,
re-create the replication agreement.
Also, the port listed for the consumer is the non-SSL port, even if the Directory Server instance is
configured to run over SSL. T his port number is used only for identification of the Directory Server
instance in the Console; it does not specify the actual port number or protocol that is used for replication.
9.13. Setting Replication Timeout Periods
Suppliers must have an exclusive connection to a consumer to send updates to the directory. As
mentioned in Section 9.5.4, “Preventing Monopolization of the Consumer in Multi-Master Replication”, it is
338
Chapter 9. Managing Replication
possible to configure a wait time for suppliers attempting to connect to a consumer, so that the supplier
does not hang while the consumer is tied up with another supplier.
It is also possible to set a timeout period for a supplier, so that it does not stay connected to a consumer
interminably attempting to send updates over a slow or broken connection.
T here are two attributes which set the timeout period:
nsDS5ReplicaTimeout sets the number of seconds that the replication operation waits for a
response from the consumer before timing out and failing. T o set the optimum number, check the
access logs to see the average amount of time that the replication process takes, and set the
timeout period accordingly.
nsDS5DebugReplicaTimeout sets the timeout period for the replication operation when debug
logging is enabled. T his setting may be appreciably higher than the nsDS5ReplicaTimeout setting
because debug logging can slow down directory operations. T his attribute can optionally set an error
log level where this parameter is applied; the default is replication debugging (8192).
NOTE
T he timeout period is limited to the maximum 32-bit integer in seconds, which translates to 24.8
days.
Both of these attributes are set in the configuration for the replicated suffix. For example, this sets
timeout periods for the ou=People suffix:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389
dn: cn=replica,cn=ou=People\,dc=example\,dc=com,cn=mapping tree,cn=config
changetype: modify
add: nsDS5ReplicaTimeout
nsDS5ReplicaTimeout: 600
add: nsDS5DebugReplicaTimeout
nsDS5DebugReplicaTimeout: 6000
9.14. Replicating o=NetscapeRoot for Admin Server Failover
Replication usually occurs between Directory Server user databases to distribute directory data, but it is
also possible to use replication to provide failover support for the Admin Server database,
o=NetscapeRoot.
1. Install and configure the first Directory Server instance.
T he setup-ds-adm in.pl script has an option, -f, which references an inf. T he inf can be
used to import LDIF files through the ConfigFile parameter, and the LDIF files can create
databases, suffixes, and replication entries. (T he inf file is described in more detail in the
Directory Server Installation Guide.)
/usr/sbin/setup-ds-admin.pl -f /tmp/server1.inf
T o configure the o=NetscapeRoot database on server1 as a multi-master supplier replica,
use the following statements in the inf file:
339
Red Hat D irectory Server 8.2 Administration Guide
[slapd]
...
ConfigFile = repluser.ldif Example 9.1, “Example Supplier Bind DN Entry”
ConfigFile = changelog.ldif Example 9.2, “Example Changelog Entry”
ConfigFile = replica.ldif Example 9.3, “Example Supplier Replica Entry”
ConfigFile = replagreement.ldif Example 9.4, “Example Replication
Agreement Entry”
...
2. Install and configure the second Directory Server instance. For the second server,
server2.exam ple.com , use the setup-ds.pl command, which installs a Directory Server
instance without installing a local Admin Server.
/usr/sbin/setup-ds.pl -f /tmp/server2.inf
With server2, use the inf file to create and configure a o=NetscapeRoot database on
server2 as a multi-master supplier replica:
[slapd]
...
ConfigFile = netscaperootdb.ldif Example 2.1, “Example Root Suffix Entry”
ConfigFile = repluser.ldif Example 9.1, “Example Supplier Bind DN Entry”
ConfigFile = changelog.ldif Example 9.2, “Example Changelog Entry”
ConfigFile = replica.ldif Example 9.3, “Example Supplier Replica Entry”
ConfigFile = replagreement.ldif Example 9.4, “Example Replication
Agreement Entry”
...
3. Initialize the o=NetscapeRoot database on server2 from server1. Add the
nsds5replicarefresh attribute to the replication agreement on server1.
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
supplier1.example.com
dn: cn=ExampleAgreement1,cn=replica,cn=o=NetscapeRoot,cn=mapping
tree,cn=config
changetype: modify
replace: nsds5beginreplicarefresh
nsds5beginreplicarefresh: start
4. Run the register-ds-adm in.pl to create a local Admin Server on server2 and switch the
configuration directory for server2 to its own o=NetscapeRoot database from server1.
/usr/sbin/register-ds-admin.pl
5. Disable the PT A Plug-in on server2 so that it does not pass bind operations for the
administrative users in its o=NetscapeRoot to server1.
See Section 1.10.1, “Enabling Plug-ins”.
9.15. Replication with Earlier Releases
T his section provides information on how to optimize replication with earlier releases of Directory Server.
34 0
Chapter 9. Managing Replication
9.15.1. Using Legacy Replication
Directory Server 8.2 can be involved in replication with earlier releases of Directory Server, providing the
following conditions are met:
Directory Server 8.2 is a consumer.
T he legacy suppliers can be Directory Server 4.0, 4.1, and 4.1x.
T he following restrictions apply:
A legacy Directory Server and Directory Server 8.2 cannot update the same replica. However, this
version of Directory Server can have different replicas, where one is supplied by a legacy Directory
Server and the other is supplied by Directory Server 8.2.
Directory Server 8.2 cannot be a supplier for other replicas.
T he main advantage of using Directory Server 8.2 as a consumer of a legacy Directory Server is to ease
the migration of a replicated environment, especially since migration is not supported from 4.x servers to
8.2. For more information on migration, see the Directory Server Installation Guide.
9.15.2. Legacy Replication and Parent Object Classes
T here is one important difference between the way that Directory Server 4.x servers handle replicated
entries and the way that Directory Server 8.2 handles replicated entries. In Directory Server 4.x, entries
could be added without specifying parent object classes, and when those entries were modified or
replicated, the server would not automatically insert those parent object classes. For example, a user
could be added with the inetorgperson object class, but not the top or person object classes:
dn: uid=jsmith,ou=People,dc=example,dc=com
objectclass: inetorgperson
uid: jsmith
cn: John Smith
sn: Smith
However, in Directory Server 8.2, the parent object classes are automatically added to the entry when
the entry is added, modified, or replicated. T his means that the entries will be slightly different on the
Directory Server 4.x supplier and the Directory Server 8.2 consumer, because the Directory Server 8.2
entry will have the parent object classes:
dn: uid=jsmith,ou=People,dc=example,dc=com
objectclass: top
objectclass: person
objectclass: inetorgperson
uid: jsmith
cn: John Smith
sn: Smith
9.15.3. Configuring Legacy Replication
T o set up legacy replication:
1. In the Directory Server Console, click the Configuration tab.
2. Select the Replication node, and click the Legacy Consum er Settings tab in the right
pane.
3. Check the Enable Legacy Consum er checkbox.
34 1
Red Hat D irectory Server 8.2 Administration Guide
T his activates the fields in the Authentication box.
4. Specify the supplier bind DN that the legacy supplier server will use to bind to the consumer.
Optionally, specify a password at least 8 characters long.
5. Click Save.
6. Now configure legacy consumer settings for each replica that will receive updates from a legacy
supplier.
a. In the navigation tree, expand the Replication node, and select a replica that will receive
updates from the legacy supplier.
b. In the Com m on Settings area, select the Enable Replica and Updatable by a
4 .x Replica checkboxes.
T hese options are the only ones required for replication to work. Optionally, specify a
replica ID. It is not necessary to specify a supplier DN because the one specified in step 4
will be used.
c. Click Save.
7. Repeat step 6 for each read-only replica that will receive updates from a legacy supplier.
8. T o complete the legacy replication setup, configure the legacy supplier to replicate to the Directory
Server 8.2 instance. For instructions on configuring a replication agreement on a 4.x Directory
Server, refer to the documentation for the legacy Directory Server.
NOTE
T he Directory Server Console will not prevent you from configuring a database as a read-write
replica and enabling legacy consumer settings. T his makes migration easier because the
Directory Server can be configured as it should be after the migration and legacy consumer
settings only have to be active for the duration of the transition.
34 2
Chapter 9. Managing Replication
9.16. Using the Retro Changelog Plug-in
T he Retro Changelog plug-in configures Directory Server to maintain a changelog that is compatible with
the changelog implemented in Directory Server 4.0, 4.1, and 4.1x. Maintaining a retro changelog is
essential to maintain a changelog for directory clients that depend on a Directory Server 4.x-style
changelog.
T o use the retro changelog plug-in, the Directory Server 8.2 instance must be configured as a singlemaster replica.
When the Directory Server is configured to maintain a retro changelog, this changelog is stored in a
separate database under a special suffix, cn=changelog.
T he retro changelog consists of a single level of entries. Each entry in the changelog has the object
class changeLogEntry and can include the attributes listed in T able 9.5, “Attributes of a Retro
Changelog Entry”.
T able 9.5. Attributes of a Retro Changelog Entry
Attribute
Definition
changeNumber
T his single-valued attribute is always present. It
contains an integer which uniquely identifies each
change. T his number is related to the order in
which the change occurred. T he higher the
number, the later the change.
targetDN
T his attribute contains the DN of the entry that
was affected by the LDAP operation. In the case
of a m odrdn operation, the targetDN attribute
contains the DN of the entry before it was
modified or moved.
changetype
Specifies the type of LDAP operation. T his
attribute can have a value of add, delete, modify,
or m odrdn.
changes
For add and modify operations, contains the
changes made to the entry in LDIF format.
newrdn
In the case of m odrdn operations, specifies the
new RDN of the entry.
deleteoldrdn
In the case of m odrdn operations, specifies
whether the old RDN was deleted.
newSuperior
In the case of m odrdn operations, specifies the
newSuperior attribute of the entry.
T his section contains information on the following retro changelog items:
Section 9.16.1, “Enabling the Retro Changelog Plug-in”
Section 9.16.2, “T rimming the Retro Changelog”
Section 9.16.3, “Searching and Modifying the Retro Changelog”
Section 9.16.4, “Retro Changelog and the Access Control Policy”
9.16.1. Enabling the Retro Changelog Plug-in
T he retro changelog plug-in configuration information is stored in the cn=Retro Changelog
34 3
Red Hat D irectory Server 8.2 Administration Guide
Plugin,cn=plugins,cn=config entry in dse.ldif. T o enable the retro changelog plug-in from
the command line:
1. Create an LDIF file that contains the following LDIF update statements:
dn: cn=Retro Changelog Plugin,cn=plugins,cn=config
changetype: modify
replace: nsslapd-pluginEnabled
nsslapd-pluginEnabled: on
2. Use the ldapm odify command to import the LDIF file into the directory.
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com -f retro.ldif
3. Restart the server.
For information on restarting the server, see Section 1.4, “Starting and Stopping Servers”.
T he retro changelog is created in the directory tree under a special suffix, cn=changelog.
T he procedure for enabling the retro changelog plug-in from Directory Server Console is the same as for
all Directory Server plug-ins. For information, see Section 1.10.1, “Enabling Plug-ins”.
9.16.2. Trimming the Retro Changelog
T he entries in the changelog can be automatically removed after a specified period of time. T o configure
the period of time after which entries are automatically deleted from the changelog, set the nsslapdchangelogmaxage configuration attribute in the cn=Retro Changelog
Plugin,cn=plugins,cn=config entry.
T he nsslapd-changelogmaxage attribute is a single-valued attribute. Its syntax is as follows:
nsslapd-changelogmaxage: Integer timeUnit
Integer is a number, and timeUnit can be s for seconds, m for minutes, h for hours, d for days, or w for
weeks.
NOTE
T here should not be a space between the Integer and timeUnit variables. T he space in the
syntax above is intended to show that the attribute value is composed of two variable parts, not
just one. For example:
nsslapd-changelogmaxage: 2d
9.16.3. Searching and Modifying the Retro Changelog
T he changelog supports search operations and is optimized for searches that include filters of the form
(& (changeNum ber>=X)(changeNum ber<=Y)).
As a general rule, do not perform add or modify operations on the retro changelog entries, although
entries can be deleted to trim the size of the changelog. Only modify the retro changelog entry to modify
34 4
Chapter 9. Managing Replication
the default access control policy.
9.16.4. Retro Changelog and the Access Control Policy
When the retro changelog is created, the following access control rules apply by default:
Read, search, and compare rights are granted to all authenticated users (userdn=anyone, not to
be confused with anonymous access where userdn=all) to the retro changelog top entry
cn=changelog.
Write and delete access are not granted, except implicitly to the Directory Manager.
Do not grant read access to anonymous users because the changelog entries can contain modifications
to sensitive information, such as passwords. Only authenticated applications and users should be
allowed to access this information.
T o modify the default access control policy which applies to the retro changelog, modify the aci attribute
of the cn=changelog entry.
9.17. Monitoring Replication Status
T he replication status can be viewed in the Directory Server Console or Red Hat Administration Express.
Section 9.17.1, “Monitoring Replication Status from the Directory Server Console”
Section 9.17.2, “Monitoring Replication Status from Administration Express”
9.17.1. Monitoring Replication Status from the Directory Server Console
1. Select the Status tab, and then, in the left navigation tree, select Replication Status.
In the right pane, a table appears that contains information about each of the replication
agreements configured for this server.
2. Click Refresh to update the contents of the tab.
T he status information displayed is described in T able 9.6, “Directory Server Console Replication
34 5
Red Hat D irectory Server 8.2 Administration Guide
Status”.
T able 9.6. Directory Server Console Replication Status
T able Header
Description
Agreement
T he name of the replication agreement.
Replica suffix
T he suffix that is replicated.
Supplier
T he supplier server in the agreement.
Consumer
T he consumer server in the agreement.
Number of changes
T he number of changes sent to this replica since
the server started.
Last replica update began
T he time when the most recent replication update
started.
Last replica update ended
T he time when the most recent replication update
ended.
Last update message
T he status for the most recent replication
updates.
Consumer initialization
T he current status on consumer initialization (in
progress or not).
Last consumer initialization update message
T he status on the last initialization of the
consumer.
Last consumer initialization began
T he time when the initialization of the consumer
replica started.
Last consumer initialization ended
T he time when the initialization of the consumer
replica ended.
9.17.2. Monitoring Replication Status from Administration Express
Although the replication status report in the Directory Server Console shows many details, it does not
show the progress of the replication. Additionally, because one report is generated per agreement, this
lists all status reports for all different agreements.
T he repl-m onitor.pl script, which is explained in detail in the Directory Server Configuration and
Command-Line Tool Reference, monitors replication status to a greater extent by providing these
functionalities:
Lists for each supplier replica on each Directory Server discovered, server URL or alias, replica ID,
replica root, and maximum change sequence number (m axcsn).
Lists corresponding to each supplier replica listed above and for each direct or indirect consumer
replicas discovered, server URL or alias, replica root, replica type, connection type of the replication
sessions, replication schedule, replication status, supplier m axcsn, and time lag between the
consumer maxcsn and the supplier maxcsn.
T he time lag field uses different colors to indicate the degree of time lag. T he threshold for each
color is configurable.
Displays the change sequence number (CSN) in human-readable format (instead of hex strings) in
the MM/DD/YYYY HH:MI Seq# SubSeq# format, where Seq# and SubSeq# are omitted if they are
zero.
Shows the output/result in the HT ML format. T he script writes the output to an HT ML file, which can
be configured to refresh itself automatically; the refresh interval is also configurable.
34 6
Chapter 9. Managing Replication
T he script is integrated into the Red Hat Administration Express, so the replication status can be viewed
in a web browser. T he Administration Express is an HT ML-based version of Red Hat Console that
provides quick access to servers running Admin Server.
T o view in-progress status of replication in Administration Express:
1. Prepare a configuration file following the guidelines provided in the "repl-monitor.pl (Monitor
replication status)" section of the Directory Server Configuration and Command-Line Tool
Reference.
2. Open the Administration Server URL in a web browser.
http://hostname:admin_port
3. Click Red Hat Adm inistration Express, and, when prompted, log in.
4. Select a supplier Directory Server instance, and click Replication Status.
T his brings up a page for specifying the runtime parameters of the replication-monitoring tool.
5. In the Configuration file field, type the path to the configuration file created in step 1, and
click OK.
T he replication-status page appears; by default, the page gets refreshed every 300 seconds.
Each table shows the status of the changes originated from a supplier replica.
T able
Description
T able Header
T he table header shows the replica ID of the
supplier replica, the replica root, and the maximum
Change State Number (CSN) on the supplier. T he
important thing is to make sure that each supplier
LDAP server has its unique replica ID. Multiple
replica roots on one LDAP server, however, could
share the same replica ID.
T able Row
Each row represents a direct or indirect
consumer of the supplier (identified in the T able
34 7
Red Hat D irectory Server 8.2 Administration Guide
Header).
Max CSN
It is the most recent CSN the consumer has
replayed that was originated from the supplier
(identified in the T able Header).
T ime Lag
It shows the time difference between the supplier
and the consumer's max CSNs for the changes
originated from the supplier (identified in the
T able Header). A consumer is in sync with its
supplier when its time lag is 0.
Last Modify T ime
It is roughly the time when the consumer's max
CSN was replayed.
Supplier
T his column lists all the suppliers of the
consumer.
Sent/Skipped
Each supplier lists roughly how many changes
originated from the supplier (identified in the
T able Header) have been replayed or skipped by
the consumer. T he numbers are kept in suppliers'
memory only. T hey will be cleared if the supplier
is restarted.
Update Status
T he number is the status code, and the string is
the implication of the status code. Watch this
column for possible deadlock if all the suppliers
complain that they cannot acquire the busy
replica. It is normal if one of the suppliers is doing
an update while the others can't acquire the busy
replica.
9.18. Solving Common Replication Conflicts
Multi-master replication uses a loose consistency replication model. T his means that the same entries
can be changed on different servers. When replication occurs between the two servers, the conflicting
changes need to be resolved. Mostly, resolution occurs automatically, based on the timestamp
associated with the change on each server. T he most recent change takes precedence.
However, there are some cases where change conflicts require manual intervention in order to reach a
resolution. Entries that have a change conflict that cannot be resolved automatically by the replication
process contain a conflict marker attribute nsds5ReplConflict. T he nsds5ReplConflict attribute is
an operational attribute which is indexed for presence and equality, so it is simple to search for entries
that contain this attribute. For example:
ldapsearch -D adminDN -w password
-b "dc=example,dc=com" "nsds5ReplConflict=*" \* nsds5ReplConflict
T he nsds5ReplConflict attribute is already indexed for presence and equality, but for performance
reasons, if there are many conflicting entries every day, index the nsds5ReplConflict attribute in other
indexes. For information on indexing, see Chapter 7, Managing Indexes.
Section 9.18.1, “Solving Naming Conflicts”
Section 9.18.2, “Solving Orphan Entry Conflicts”
Section 9.18.3, “Solving Potential Interoperability Problems”
34 8
Chapter 9. Managing Replication
9.18.1. Solving Naming Conflicts
When two entries are created with the same DN on different servers, the automatic conflict resolution
procedure during replication renames the last entry created, including the entry's unique identifier in the
DN. Every directory entry includes a unique identifier given by the operational attribute nsuniqueid.
When a naming conflict occurs, this unique ID is appended to the non-unique DN.
For example, the entry uid=adam ss,ou=people,dc=exam ple,dc=com is created on Server A at
time t1 and on Server B at time t2, where t2 is greater (or later) than t1. After replication, Server A and
Server B both hold the following entries:
uid=adam ss,ou=people,dc=exam ple,dc=com (created at time t1)
nsuniqueid=664 4 6001-1dd211b2+uid=adam ss,dc=exam ple,dc=com (created at time t2)
T he second entry needs to be renamed in such a way that it has a unique DN. T he renaming procedure
depends on whether the naming attribute is single-valued or multi-valued.
9.18.1.1. Renaming an Entry with a Multi-Valued Naming Attribute
T o rename an entry that has a multi-valued naming attribute:
1. Rename the entry using a new value for the naming attribute, and keep the old RDN. For example:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: nsuniqueid=66446001-1dd211b2+uid=adamss,dc=example,dc=com
changetype: modrdn
newrdn: uid=NewValue
deleteoldrdn: 0
2. Remove the old RDN value of the naming attribute and the conflict marker attribute. For example:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: uid=NewValue,dc=example,dc=com
changetype: modify
delete: uid
uid: adamss
delete: nsds5ReplConflict
-
NOTE
T he unique identifier attribute nsuniqueid cannot be deleted.
T he Console does not support editing multi-valued RDNs. For example, if there are two servers in a
multi-master mode, an entry can be created on each server with the same user ID, and then the new
entries' RDN changed to the nsuniqueid uid value. Attempting to modify this entry from the Console
returns the error Changes cannot be saved for entries with multi-valued RDNs.
Opening the entry in the advanced mode shows that the naming attribute has been set to nsuniqueid
uid. However, the entry cannot be changed or corrected by changing the user ID and RDN values to
something different. For example, if jdoe was the user ID and it should be changed to jdoe1, it cannot
be done from the Console. Instead, use the ldapm odify command:
34 9
Red Hat D irectory Server 8.2 Administration Guide
be done from the Console. Instead, use the ldapm odify command:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=John Doe
changetype: modify
replace: uid
uid: jdoe
dn: cn=John Doe
changetype: modrdn
newrdn: uid=jdoe1
deleteoldrdn: 1
9.18.1.2. Renaming an Entry with a Single-Valued Naming Attribute
T o rename an entry that has a single-valued naming attribute:
1. Rename the entry using a different naming attribute, and keep the old RDN. For example:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: nsuniqueid=66446001-1dd211b2+dc=pubs,dc=example,dc=com
changetype: modrdn
newrdn: cn=TempValue
deleteoldrdn: 0
2. Remove the old RDN value of the naming attribute and the conflict marker attribute. For example:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=TempValue,dc=example,dc=com
changetype: modify
delete: dc
dc: pubs
delete: nsds5ReplConflict
-
NOTE
T he unique identifier attribute nsuniqueid cannot be deleted.
3. Rename the entry with the intended attribute-value pair. For example:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=TempValue,dc=example,dc=com
changetype: modrdn
newrdn: dc=NewValue
deleteoldrdn: 1
Setting the value of the deleteoldrdn attribute to 1 deletes the temporary attribute-value pair
cn=TempValue. T o keep this attribute, set the value of the deleteoldrdn attribute to 0.
350
Chapter 9. Managing Replication
9.18.2. Solving Orphan Entry Conflicts
When a delete operation is replicated and the consumer server finds that the entry to be deleted has
child entries, the conflict resolution procedure creates a glue entry to avoid having orphaned entries in
the directory.
In the same way, when an add operation is replicated and the consumer server cannot find the parent
entry, the conflict resolution procedure creates a glue entry representing the parent so that the new
entry is not an orphan entry.
Glue entries are temporary entries that include the object classes glue and extensibleObject. Glue
entries can be created in several ways:
If the conflict resolution procedure finds a deleted entry with a matching unique identifier, the glue
entry is a resurrection of that entry, with the addition of the glue object class and the
nsds5ReplConflict attribute.
In such cases, either modify the glue entry to remove the glue object class and the
nsds5ReplConflict attribute to keep the entry as a normal entry or delete the glue entry and its
child entries.
T he server creates a minimalistic entry with the glue and extensibleObject object classes.
In such cases, modify the entry to turn it into a meaningful entry or delete it and all of its child entries.
9.18.3. Solving Potential Interoperability Problems
For reasons of interoperability with applications that rely on attribute uniqueness, such as a mail server,
it may be necessary to restrict access to the entries which contain the nsds5ReplConflict attribute. If
access is not restricted to these entries, then the applications requiring one attribute only pick up both
the original entry and the conflict resolution entry containing the nsds5ReplConflict, and operations
will fail.
T o restrict access, modify the default ACI that grants anonymous read access:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: dc=example,dc=com
changetype: modify
delete: aci
aci: (target ="ldap:///dc=example,dc=com")(targetattr
!="userPassword")(version 3.0;acl "Anonymous read-search
access";allow (read, search, compare)(userdn = "ldap:///anyone");)
add: aci
aci: (target="ldap:///dc=example,dc=com")(targetattr!="userPassword")
(targetfilter="(!(nsds5ReplConflict=*))")(version 3.0;acl
"Anonymous read-search access";allow (read, search, compare)
(userdn="ldap:///anyone");)
-
T he new ACI filters out all entries that contain the nsds5ReplConflict attribute from search results.
9.19. Troubleshooting Replication-Related Problems
T his section lists some error messages, explains possible causes, and offers remedies.
351
Red Hat D irectory Server 8.2 Administration Guide
It is possible to get more debugging information for replication by setting the error log level to 8192,
which is replication debugging. See Section 15.3.5, “Configuring Log Levels”.
T o change the error log level to 8192 with ldapm odify:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=config
changetype: modify
replace: nsslapd-errorlog-level
nsslapd-errorlog-level: 8192
Because log level is additive, running the above command will result in excessive messages in the error
log. So, use it judiciously.
T o turn off replication debugging log, set the same attribute to 0.
T he cl-dum p.pl script, which is explained in detail in the Directory Server Configuration and
Command-Line Tool Reference can also help troubleshoot replication-related problems. Depending on
the usage options, the script can selectively dump a particular replica:
Dump the contents of a replication-change-log file and in-memory variables purge RUV and
m axRUV.
Grep and interpret change sequence numbers (CSNs) in the changelog.
Get the base-64 encoded changelog from the Directory Server, and then decode the changelog.
Many common replication problems are described in T able 9.7, “Replication Errors”.
352
Chapter 9. Managing Replication
T able 9.7. Replication Errors
Error/Symptom
Reason
Impact
Remedy
agmt=%s (%s:%d)
Replica has a different
generation ID than the
local data
T he consumer
specified at the
beginning of this
message has not been
(successfully) initialized
yet, or it was initialized
from a different root
supplier.
T he local supplier will
not replicate any data
to the consumer.
Ignore this message if
it occurs before the
consumer is initialized.
Otherwise, reinitialize
the consumer if the
message is persistent.
In a multi-master
environment, all the
servers should be
initialized only once
from a root supplier,
directly or indirectly. For
example, M1 initializes
M2 and M4, M2 then
initializes M3, and so
on. T he important thing
to note is that M2 must
not start initializing M3
until M2's own
initialization is done
(check the total update
status from the M1's
Console or M1 or M2's
error log). Also, M2
should not initialize M1
back.
Warning: data for
replica's was reloaded,
and it no longer
matches the data in the
changelog. Recreating
the changelog file. T his
could affect replication
with replica's
consumers, in which
case the consumers
should be reinitialized.
T his message may
appear only when a
supplier is restarted. It
indicates that the
supplier was unable to
write the changelog or
did not flush out its
RUV at its last
shutdown. T he former
is usually because of a
disk-space problem,
and the latter because
a server crashed or
was ungracefully shut
down.
T he server will not be
able to send the
changes to a consumer
if the consumer's
maxcsn no longer
exists in the server's
changelog.
Check the disk space
and the possible core
file (under the server's
logs directory). If this is
a single-master
replication, reinitialize
the consumers.
Otherwise, if the server
later complains that it
can't locate some CSN
for a consumer, see if
the consumer can get
the CSN from other
suppliers. If not,
reinitialize the
consumer.
agmt=%s(%s:%d):
Can't locate CSN %s in
the changelog (DB
rc=%d). T he consumer
may need to be
reinitialized.
Most likely the
changelog was
recreated because of
the disk is full or the
server ungracefully
shutdown.
T he local server will not
be able to send any
more change to that
consumer until the
consumer is
reinitialized or gets the
CSN from other
suppliers.
If this is a singlemaster replication,
reinitialize the
consumers. Otherwise,
see if the consumer
can get the CSN from
other suppliers. If not,
reinitialize the
353
Red Hat D irectory Server 8.2 Administration Guide
consumer.
T oo much time skew
T he system clocks on
the host machines are
extremely out of sync.
T he system clock is
used to generate a part
of the CSN. In order to
reflect the change
sequence among
multiple suppliers,
suppliers would
forward-adjust their
local clocks based on
the remote clocks of the
other suppliers.
Because the
adjustment is limited to
a certain amount, any
difference that exceeds
the permitted limit will
cause the replication
session to be aborted.
Synchronize the
system clocks on the
Directory Server host
machines. If applicable,
run the network time
protocol (ntp) daemon
on those hosts.
agmt=%s(%s:%d):
Warning: Unable to
send endReplication
extended operation
(%s)
T he consumer is not
responding.
If the consumer
recovers without being
restarted, there is a
chance that the replica
on the consumer will be
locked forever if it did
not receive the release
lock message from the
supplier.
Watch if the consumer
can receive any new
change from any of its
suppliers, or start the
replication monitor, and
see if all the suppliers
of this consumer warn
that the replica is busy.
If the replica appears to
be locked forever and
no supplier can get in,
restart the consumer.
Changelog is getting
too big.
Either changelog purge
is turned off, which is
the default setting, or
changelog purge is
turned on, but some
consumers are way
behind the supplier.
354
By default changelog
purge is turned off. T o
turn it on from the
command line, run
ldapm odify as
follows:
Chapter 9. Managing Replication
behind the supplier.
/usr/lib64/mozl
dap/ldapmodify
-D
"cn=directory
manager" -w
secret -p 389 -h
server.example.c
om
dn:
cn=changelog5,cn
=config
changetype:
modify
add: nsslapdchangelogmaxage
nsslapdchangelogmaxage:
1d
where 1d means 1 day.
Other valid time units
are s for seconds, m for
minutes, h for hours,
and w for weeks. A
value of 0 turns off the
purge.
With changelog purge
turned on, a purge
thread that wakes up
every five minutes will
remove a change if its
age is greater than the
value of nsslapdchangelogmaxage and
if it has been replayed
to all the direct
consumers of this
supplier (supplier or
hub).
If it appears that the
changelog is not
purged when the purge
threshold is reached,
check the maximum
time lag from the
replication monitor
among all the
consumers.
Irrespective of what the
purge threshold is, no
change will be purged
before it is replayed by
355
Red Hat D irectory Server 8.2 Administration Guide
all the consumers.
T he Replication Monitor
is not responding. (For
information on
Replication Monitor,
see Section 9.17,
“Monitoring Replication
Status”.)
In the Replication
Monitor, some
consumers show just
the header of the table.
(For information on
Replication Monitor,
see Section 9.17,
“Monitoring Replication
Status”.)
356
T he SSL port is
specified in some
replication agreement,
but the certificate
database is not
specified or not
accessible by the
Replication Monitor. If
there is no SSL port
problem, one of the
servers in the
replication topology
might hang.
Map the SSL port to a
non-SSL port in the
configuration file of the
Replication Monitor. For
example, if 636 is the
SSL port and 389 is the
non-SSL port, add the
following line in the
[connection]
section:
No change has
originated from the
corresponding
suppliers. In this case,
the MaxCSN: in the
header part should be
"None".
T here is nothing wrong
if there is no change
originated from a
supplier.
*:636=389:*:pass
word
Chapter 10. Synchronizing Red Hat D irectory Server with Microsoft Active D irectory
Chapter 10. Synchronizing Red Hat Directory Server with
Microsoft Active Directory
Windows Sync carries over changes in a directory — adds, deletes, and changes in groups, users, and
passwords — between Red Hat Directory Server and Microsoft Active Directory. T his makes it much
more efficient and effective to maintain consistent information across directories.
10.1. About Windows Sync
Synchronization allows the user and group entries in Active Directory to be matched with the entries in
the Red Hat Directory Server. As entries are created, modified, or deleted, the corresponding change is
made to the sync peer server, allowing two-way synchronization of users, passwords, and groups.
T he synchronization process is analogous to the replication process: the synchronization is enabled by
a plug-in, configured and initiated through a sync agreement, and record of directory changes is
maintained and updates are sent according to that changelog. T his synchronizes users and groups
between Directory Server and a Windows server.
Windows Sync has two parts is configured in two parts, one for user and group entries and the other for
passwords:
Directory Server Windows Sync. Synchronization for user and group entries is configured in a
synchronization agreement, much like replication is configured in a replication agreement. A sync
agreement defines what kinds of entries are synchronized (users, groups, or both) and which
direction changes are synced (from the Directory Server to Active Directory, from Active Directory to
Directory Server, or both).
T he Directory Server relies on the Multi-Master Replication Plug-in to synchronize user and group
entries. T he same changelog that is used for multi-master replication is also used to send updates
from the Directory Server to Active Directory as LDAP operations. T he server also performs LDAP
search operations against its Windows server to synchronize changes made to Windows entries to
the corresponding Directory Server entry.
Password Sync Service. Password changes made on Directory Server are automatically synced over
to Active Directory, but there must be a special hook to recognize and transmit password changes on
Active Directory over to Directory Server. T his is done by the Password Sync Service. T his
application captures password changes on the Windows machines and send them to the Directory
Server over LDAPS.
T he Password Sync Service must be installed on every Active Directory domain controller.
Figure 10.1. Active Directory — Directory Server Synchronization Process
357
Red Hat D irectory Server 8.2 Administration Guide
Synchronization is configured and controlled by one or more synchronization agreements, which
establishes synchronization between sync peers, the directory servers being synced. T hese are similar
in purpose to replication agreements and contain a similar set of information, including the hostname and
port number for Active Directory. T he Directory Server connects to its peer Windows server via
LDAP/LDAPS to both send and receive updates.
LDAP, a standard connection, can be used for syncing user and group entries alone, but to synchronize
passwords, some sort of secure connection is required. If a secure connection is not used, the Windows
domain will not accept password changes from the Directory Server and the Password Sync Service will
not send passwords from the Active Directory domain to the Directory Server. Win Sync allows both
LDAPS using SSL/T LS and Start T LS.
A single Active Directory subtree is synchronized with a single Directory Server subtree, and vice versa.
Unlike replication, which connects databases, synchronization is between suffixes, parts of the directory
tree structure. T he synced Active Directory and Directory Server suffixes are both specified in the sync
agreement. All entries within the respective subtrees are candidates for synchronization, including
entries that are not immediate children of the specified suffix DN.
NOTE
Any descendant container entries need to be created separately in Active Directory by an
administrator; Windows Sync does not create container entries.
T he Directory Server maintains a changelog, a database that records modifications that have occurred.
T he changelog is used by Windows Sync to coordinate and send changes made to the Active Directory
peer. Changes to entries in Active Directory are found by using Active Directory's Dirsync search
feature. T he Dirsync search is issued periodically, every five minutes, to check for changes on the Active
Directory server. Using Dirsync ensures that only those entries that have changed since the previous
search are retrieved.
In some situations, such as when synchronization is configured or there have been major changes to
directory data, a total update, or resynchronization, can be run. T his examines every entry in both sync
peers and sends any modifications or missing entries. A full Dirsync search is initiated whenever a total
update is run. See Section 10.6, “Sending Synchronization Updates” for more information.
Windows Sync provides some control over which entries are synchronized to grant administrators finegrained control of the entries that are synchronized and to give sufficient flexibility to support different
deployment scenarios. T his control is set through different configuration attributes set in the Directory
Server:
When creating the sync agreement, there is an option to synchronizing new Windows entries
(nsDS7NewWinUserSyncEnabled and nsDS7NewWinGroupSyncEnabled) as they are created. If
these attributes are set to on, then existing Windows users/groups are synchronized to the
Directory Server, and users/groups as they are created are synchronized to the Directory Server.
Within the Windows subtree, only entries with user or group object classes can be synchronized to
Directory Server.
On the Directory Server, only entries with the ntUser or ntGroup object classes and attributes can
be synchronized.
T he placement of the sync agreement depends on what suffixes are synchronized; for a single suffix,
the sync agreement is made for that suffix alone; for multiple suffixes, the sync agreement is made at a
higher branch of the directory tree. T o propagate Windows entries and updates throughout the Directory
Server deployment, make the agreement between a master in a multi-master replication environment,
358
Chapter 10. Synchronizing Red Hat D irectory Server with Microsoft Active D irectory
and use that master to replicate the changes across the Directory Server deployment, as shown in
Figure 10.2, “Multi-Master Directory Server — Windows Domain Synchronization”.
IMPORTANT
You cannot configure a sync agreement on a hub. Synchronization will not succeed.
WARNING
T here can only be a single sync agreement between the Directory Server environment and the
Active Directory environment. Multiple sync agreements to the same Active Directory domain can
create entry conflicts.
Figure 10.2. Multi-Master Directory Server — Windows Domain Synchronization
Directory Server passwords are synchronized along with other entry attributes because plain-text
passwords are retained in the Directory Server changelog. T he Password Sync service is needed to
catch password changes made on Active Directory. Without the Password Sync service, it would be
impossible to have Windows passwords synchronized because passwords are hashed in Active
Directory, and the Windows hashing function is incompatible with the one used by Directory Server.
10.2. Configuring Windows Sync
Configuring synchronization is very similar to configuring replication. It requires configuring the database
as a master with a changelog and creating an agreement to define synchronization. A common user
identity, a sync user, connects to the Windows sync peer to send updates from the Directory Server and
to check for updates to sync back to the Directory Server.
359
Red Hat D irectory Server 8.2 Administration Guide
NOTE
T o synchronize passwords (which is the only way for users to be active on both Directory Server
and Active Directory), synchronization must be configured to run over SSL/T LS. T herefore, this
configuration section assumes that SSL/T LS must also be configured.
Configuring synchronization over SSL/T LS is also similar to configuring replication over SSL/T LS. Both
sync peers must be configured to trust each other for encrypted sessions (all password operations are
performed over T LS/SSL).
All synchronization for user and group entries is passive from the Active Directory side; it is the Directory
Server which sends updates on its side and polls for updates on the Active Directory domain. For
passwords, the Active Directory server requires a separate password service; this service actively
sends password changes from the Active Directory domain to Directory Server.
10.2.1. Step 1: Configure SSL on Directory Server
T he full instructions for configuring the Directory Server to run in SSL are at Section 14.2.3.1, “Enabling
T LS/SSL Only in the Directory Server”. Basically, the Directory Server needs to have the appropriate SSL
certificates installed, be configured to run over an SSL port, and allow client authentication from other
servers.
T wo certificates must be issued and installed on both the Directory Server and the Active Directory sync
peer:
CA certificate, shared between the Directory Server and Active Directory
Server certificates for the Directory Server and Active Directory sync peers, which are accessible by
the sync services
T o set up SSL:
1. Generate a certificate request.
a. In the Directory Server Console, select the T asks tab, and click Manage Certificates.
b. Select the Server Certs tab, and click the Request button at the bottom.
c. Fill in the certificate information, and save the certificate request to a file.
2. Submit the certificate to a certificate authority, and retrieve it once it is issued.
T he method for submitting certificate requests and retrieving certificates varies for each CA.
360
Chapter 10. Synchronizing Red Hat D irectory Server with Microsoft Active D irectory
3. Install the new certificate.
a. In the Directory Server Console, select the T asks tab, and click Manage Certificates.
b. Select the Server Certs tab, and click Install at the bottom of the window.
c. Paste in the certificate, and set the password for the token database.
4. Install the CA certificate for the issuing CA.
a. Download and save the CA certificate from the CA's site. Each CA has a slightly different
way of making its CA certificate available.
b. In the Directory Server Console, select the T asks tab, and click Manage Certificates.
c. Go to the CA Certs tab, and click Install at the bottom of the window.
d. Paste in the CA certificate or point to the downloaded file, and go through the certificate
installer.
5. Change the server to the SSL port; this is described in much more detail in Section 1.7, “Changing
Directory Server Port Numbers”.
a. Open the Directory Server Console, and open the Configuration tab for the Directory
Server.
b. In the Settings tab, set the secure port for the server to use for T LS/SSL communications,
such as 636. Click Save.
361
Red Hat D irectory Server 8.2 Administration Guide
c. Select the Encryption tab in the right pane.
d. Select the Enable SSL for this Server checkbox, then select the certificate to use
from the drop-down menu. Click Save.
e. Restart the Directory Server. T he Directory Server must be restarted from the command
line.
service dirsrv restart example
T o restart the Directory Server without the password prompt, create a PIN file or use a
hardware crypto device. See Section 14.2.3.3, “Creating a Password File for the Directory
Server” for information on how to create a PIN file.
10.2.2. Step 2: Configure the Active Directory Domain
NOTE
Synchronization can only be configured with an Active Directory domain controller, so make sure
that the domain is properly installed and configured.
T he first configuration step is to make sure that the Active Directory password complexity policies are
are enabled so that the Password Sync service will run.
1. Run secpol.m sc from the command line.
2. Select Security Settings.
3. Open Account Policies, and then open Password Policy.
362
Chapter 10. Synchronizing Red Hat D irectory Server with Microsoft Active D irectory
4. Enable the Password m ust m eet com plexity requirem ents option and save.
If SSL is not already enabled, set up SSL on the Active Directory server. Setting up LDAPS is explained in
more detail in the Microsoft knowledgebase at http://support.microsoft.com/kb/321051.
1. Install a certificate authority in the Windows Com ponents section in Add/Rem ove Program s.
2. Select the Enterprise Root CA option.
3. Reboot the Active Directory server. If IIS web services are running, the CA certificate can be
accessed by opening http://servername/certsrv.
4. Set up the Active Directory server to use the SSL server certificate.
a. Create a certificate request .inf, using the fully-qualified domain name of the Active
Directory as the certificate subject. For example:
;----------------- request.inf ----------------[Version]
Signature="$Windows NT$
[NewRequest]
Subject = "CN=ad.server.example.com, O=Engineering, L=Raleigh, S=North
Carolina, C=US"
KeySpec = 1
KeyLength = 2048
Exportable = TRUE
MachineKeySet = TRUE
SMIME = False
PrivateKeyArchive = FALSE
UserProtected = FALSE
UseExistingKeySet = FALSE
ProviderName = "Microsoft RSA SChannel Cryptographic Provider"
ProviderType = 12
RequestType = PKCS10
KeyUsage = 0xa0
[EnhancedKeyUsageExtension]
OID=1.3.6.1.5.5.7.3.1
;-----------------------------------------------
For more information on the .inf request file, see the Microsoft documentation, such as
http://technet.microsoft.com/en-us/library/cc783835.aspx.
363
Red Hat D irectory Server 8.2 Administration Guide
b. Generate the certificate request.
certreq -new request.inf request.req
c. Submit the request to the Active Directory CA. For example:
certreq -submit request.req certnew.cer
NOTE
If the command-line tool returns an error message, then use the Web browser to
access the CA and submit the certificate request. If IIS is running, then the CA URL is
http://servername/certsrv.
d. Accept the certificate request. For example:
certreq -accept certnew.cer
e. Make sure that the server certificate is present on the Active Directory server.
In the File menu, click Add/Remove, then click Certificates and Personal>Certificates.
f. Import the CA certificate from Directory Server into Active Directory. Click T rusted Root
CA, then Im port, and browse for the Directory Server CA certificate.
5. Reboot the domain controller.
T o test that the server is running in SSL correctly, try searching the Active Directory over LDAPS.
10.2.3. Step 3: Select or Create the Sync Identity
T here are two users used to configure Windows Sync:
An Active Directory user, specified in the sync agreement.
T he user specified in the sync agreement is the entity as whom the Directory Server binds to Active
Directory to send and receive updates. T he Active Directory user should be a member of the Domain
Admins group, or have equivalent rights, and must have rights to replicate directory changes.
For information on adding users and setting privileges in Active Directory, see the Microsoft
documentation.
A Directory Server user, specified in the Password Sync Service.
T he user referenced in the Password Sync Service must have read and write permissions to every
entry within the synchronized subtree and absolutely must have write access to password attributes
in Directory Server so that Password Sync can update password changes.
NOTE
T he user cited in the sync agreement (the supplier DN) exists on the Active Directory server. T he
user cited in the Password Sync configuration exists on Directory Server.
T o create a sync user on Directory Server:
1. Create a new entry, such as cn=sync user,cn=config, with a password. For example:
364
Chapter 10. Synchronizing Red Hat D irectory Server with Microsoft Active D irectory
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389
-h server.example.com
dn: cn=sync user,cn=config
objectClass: inetorgperson
objectClass: person
objectClass: top
cn: sync user
sn: SU
userPassword: secret
passwordExpirationTime: 20380119031407Z
2. Set an ACI that grants the sync user access to compare and write user passwords.
T he ACI must be set at the top of the subtree which will be synchronized. For example:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389
dn: ou=people,dc=example,dc=com
changetype: modify
add: aci
aci: (targetattr="userPassword")(version 3.0;aci "password sync";allow
(write,compare) userdn="ldap:///cn=sync user,cn=config";)
For security reasons, the Password Sync user should not be Directory Manager and should not
be part of the synchronized subtree.
10.2.4. Step 4: Install the Password Sync Service
Password Sync can be installed on every domain controller in the Active Directory domain in order to
synchronize Windows passwords.
Passwords can only be synchronized if both the Directory Server and Windows server are running in
SSL, the sync agreement is configured over an SSL connection, and certificate databases are configured
for Password Sync to access.
1. Download the PassSync.m si file from the appropriate Directory Server channel in Red Hat
Network and save it to the Active Directory machine.
NOTE
T here are two PassSync packages available, one for 32-bit Windows servers and one for
64-bit. Make sure to select the appropriate packages for your Windows platform.
2. Double-click on the PassSync.m si file to install it.
3. T he Password Sync Setup window appears. Hit Next to begin installing.
4. Fill in the Directory Server hostname, secure port number, user name (such as cn=sync
user,cn=config), the certificate token (password), and the search base (e.g.,
ou=People,dc=exam ple,dc=com ).
365
Red Hat D irectory Server 8.2 Administration Guide
Hit Next, then Finish to install Password Sync.
5. Reboot the Windows machine to start Password Sync.
NOTE
T he Windows machine must be rebooted. Without the rebooting, PasswordHook.dll is
not enabled, and password synchronization will not function.
T he first attempt to synchronize passwords, which happened when the Password Sync application is
installed, will always fail because the SSL connection between the Directory Server and Active Directory
sync peers. T he tools to create the certificate and key databases is installed with the .m si.
Password Sync and many of its libraries are installed in C:\Program Files\Red Hat Directory
Password Synchronization. All of the files installed with Password Sync are listed in T able 10.1,
“Installed Password Sync Libraries”.
366
Chapter 10. Synchronizing Red Hat D irectory Server with Microsoft Active D irectory
T able 10.1. Installed Password Sync Libraries
Directory
Library
Directory
Library
C:\WINDOWS\system3
2
passhook.dll
C:\WINDOWS\system32
libnspr4.dll
C:\WINDOWS\system3
2
nss3.dll
C:\WINDOWS\system32
sqlite3.dll
C:\WINDOWS\system3
2
softokn3.dll
C:\WINDOWS\system32
nssdbm3.dll
C:\WINDOWS\system3
2
nssutil3.dll
C:\WINDOWS\system3
2
smime3.dll
C:\WINDOWS\system32
freebl3.dll
C:\Program Files\Red
Hat Directory Password
Synchronization
nsldap32v60.dll
C:\Program Files\Red
Hat Directory Password
Synchronization
certutil.exe
C:\Program Files\Red
Hat Directory Password
Synchronization
nsldappr32v60.dll
C:\Program Files\Red
Hat Directory Password
Synchronization
nsldapssl32v60.dll
C:\WINDOWS\system3
2
ssl3.dll
C:\WINDOWS\system32
libplc4.dll
C:\Program Files\Red
Hat Directory Password
Synchronization
nssckbi.dll
C:\Program Files\Red
Hat Directory Password
Synchronization
nsldif32v60.dll
C:\Program Files\Red
Hat Directory Password
Synchronization
passsync.log
C:\Program Files\Red
Hat Directory Password
Synchronization
passsync.exe
C:\Program Files\Red
Hat Directory Password
Synchronization
pk12util.exe
C:\Program Files\Red
Hat Directory Password
Synchronization
msvcr71.dll
C:\WINDOWS\system3
2
libplds4.dll
[a]
[a] This lo g file is no t an ins talled lib rary, b ut it is c reated at ins tallatio n.
10.2.5. Step 5: Configure the Password Sync Service
Next, set up certificates that Password Sync uses to access the Directory Server over SSL:
NOTE
SSL is required for Password Sync to send passwords to Directory Server. T he service will not
send the passwords except over SSL to protect the clear text password sent from the Active
Directory machine to the Directory Server machine. T his means that Password Sync will not work
until SSL is configured.
1. On the Directory Server, export the server certificate.
367
Red Hat D irectory Server 8.2 Administration Guide
cd /etc/dirsrv/slapd-instance_name
certutil -d . -L -n "CA certificate" -a > dsca.crt
2. Copy the exported certificate from the Directory Server to the Windows machine.
3. Open a command prompt on the Windows machine, and open the Password Sync installation
directory.
cd "C:\Program Files\Red Hat Directory Password Synchronization"
4. Create new cert8.db and key.db databases on the Windows machine.
certutil.exe -d . -N
5. Import the server certificate from the Directory Server into the new certificate database.
certutil.exe -d . -A -n "DS CA cert" -t CT,, -a -i \path\to\dsca.crt
6. Verify that the CA certificate was correctly imported.
certutil.exe -d . -L -n "DS CA cert"
7. Reboot the Windows machine. T he Password Sync service is not available until after a system
reboot.
NOTE
If any Active Directory user accounts exist when Password Sync is first installed, then the
passwords for those user accounts cannot be synchronized until they are changed because
Password Sync cannot decrypt a password once it has been hashed in Active Directory.
10.2.6. Step 6: Configure the Directory Server Database for Synchronization
Just as with replication, there must be a changelog available to track and send directory changes and
the Directory Server database being synchronized must be configured as a replica.
NOTE
If the Directory Server database is already configured for replication, this step is not necessary.
Setting up a database for replication is described in Section 9.5.1, “Configuring the Read-Write Replicas
on the Supplier Servers”.
10.2.6.1. Setting up the Directory Server from the Console
First, enable the changelog:
1. In the Directory Server Console, select the Configuration tab.
2. In the left-hand navigation tree, click the Replication folder.
3. In the main window, click the Supplier Settings tab.
4. Check the Enable Changelog database.
368
Chapter 10. Synchronizing Red Hat D irectory Server with Microsoft Active D irectory
5. Set the changelog database directory. Click the Use default button to use the default or
Browse... to select a custom directory.
6. Save the changelog settings.
After setting up the changelog, then configure the database that will be synchronized as a replica. T he
replica role should be either a single-master or multi-master.
IMPORTANT
You cannot configure a sync agreement on a hub. Synchronization will not succeed.
1. In the Directory Server Console, select the Configuration tab.
2. In the left-hand navigation tree, click the Replication folder, then click the name of the
database to synchronize.
By default, there are two databases, NetscapeRoot for directory configuration and userRoot
for directory entries. Other databases may be listed if they have been added to Directory Server.
3. Check the Enable Replica checkbox, and select the radio button by the type of replica which
the database is.
369
Red Hat D irectory Server 8.2 Administration Guide
4. In the Update Settings section, either select or add a supplier DN. T his is the user account as
which synchronization process will be run. As mentioned in Section 10.2.3, “Step 3: Select or
Create the Sync Identity”, this user must be on the Active Directory server.
5. Save the replication settings for the database.
NOTE
For more information on replication settings, see Chapter 9, Managing Replication.
10.2.6.2. Setting up the Directory Server for Sync from the Command Line
First, enable the changelog:
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=changelog5,cn=config
objectclass: top
objectclass: extensibleObject
cn: changelog5
nsslapd-changelogdir: /var/lib/dirsrv/slapd-instance_name/changelogdb
T hen, create the supplier replica entry:
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=sync replica,cn=dc=example\,dc=com,cn=mapping tree,cn=config
objectclass: top
objectclass: nsds5replica
objectclass: extensibleObject
cn: sync replica
nsds5replicaroot: dc=example,dc=com
nsds5replicaid: 7
nsds5replicatype: 3
nsds5flags: 1
nsds5ReplicaPurgeDelay: 604800
nsds5ReplicaBindDN: cn=sync user,cn=config
370
Chapter 10. Synchronizing Red Hat D irectory Server with Microsoft Active D irectory
T hese different parameters are described in more detail in the Configuration and Command-Line Tool
Reference and Section 9.7.1, “Configuring Suppliers from the Command Line”.
10.2.7. Step 7: Create the Synchronization Agreement
Create the synchronization agreement.
NOTE
If secure binds are required for simple password authentication (Section 13.5.1, “Requiring
Secure Binds”), then any replication operations will fail unless they occur over a secure
connection. Using a secure connection (SSL/T LS and Start T LS connections or SASL
authentication) is recommended, anyway.
10.2.7.1. Creating the Sync Agreement from the Console
1. In the Directory Server Console, select the Configuration tab.
2. In the left-hand navigation tree, click Replication, then right-click on the database to sync. T he
default user database is userRoot, but additional databases are added as new suffixes are
added to the Directory Server.
Alternatively, highlight the database, and in the top tool bar, click Object.
3. Select New Windows Sync Agreem ent from the menu.
4. In the two fields, supply a name and description of the synchronization agreement. Hit Next.
5. In the Windows Sync Server Info window, fill in the Active Directory information in the
Windows Dom ain Inform ation area.
371
Red Hat D irectory Server 8.2 Administration Guide
T he name of the Windows domain.
What kinds of entries to synchronize; users and groups are synchronized independently.
When a type of entry is chosen, then all of the entries of that type that are found in the
Windows subtree are created in the Directory Server.
T he Windows and Directory Server subtree information; this is automatically filled in.
T he hostname of the domain controller
T he Windows server's port number
6. Set the connection type. T here are three options:
Use LDAP. T his sets either a standard, unencrypted connection.
Use TLS/SSL. T his uses a secure connection over the server's secure LDAPS port, such as
636. Both the Directory Server and the Windows server must be properly configured to run in
T LS/SSL for this connection and must have installed each other's CA certificates in order to
trust their server certificates.
Use Start TLS. T his uses Start T LS to establish a secure connection over the server's
standard port. Like regular SSL, these peer servers must be able to trust each other's
certificates.
Using either T LS/SSL or Start T LS is recommended for security reasons. T LS/SSL or Start T LS is
required for synchronizing passwords because Active Directory refuses to modify passwords
372
Chapter 10. Synchronizing Red Hat D irectory Server with Microsoft Active D irectory
unless the connection is SSL-protected.
7. Fill in the authentication information in the Bind as... and Password fields with the sync ID
information. T his user must exist in the Active Directory domain.
8. Save the sync agreement.
NOTE
By default, Win Sync polls the Active Directory peer every five (5) minutes to check for changes. In
the sync agreement summary, this is displayed as the Update Interval. T he update interval
can be changed by editing the winSyncInterval attribute manually. See Section 10.7, “Modifying
the Sync Agreement”.
When the agreement is complete, the new sync agreement is listed under the suffix.
10.2.7.2. Creating the Sync Agreement from the Command Line
It is also possible to add the sync agreement through the command line.
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=ExampleSyncAgreement,cn=sync replica,cn=dc=example\,dc=com,cn=mapping
tree,cn=config
changetype: add
objectclass: top
objectclass: nsDSWindowsReplicationAgreement
cn: ExampleSyncAgreement
nsds7WindowsReplicaSubtree: cn=Users,dc=ad1
nsds7DirectoryReplicaSubtree: ou=People,dc=example,dc=com
nsds7NewWinUserSyncEnabled: on
nsds7NewWinGroupSyncEnabled: on
nsds7WindowsDomain: ad1
nsDS5ReplicaRoot: dc=example,dc=com
nsDS5ReplicaHost: ad1.windows-server.com
nsDS5ReplicaPort: 389
nsDS5ReplicaBindDN: cn=sync user,cn=config
nsDS5ReplicaBindCredentials: {DES}ffGad646dT0nnsT8nJOaMA==
nsDS5ReplicaTransportInfo: TLS
winSyncInterval: 1200
All of the different parameters used in the sync agreement are listed in T able 10.6, “Sync Agreement
Attributes”. T hese different parameters are described in more detail in the Configuration and CommandLine Tool Reference.
10.2.8. Step 8: Configure Directory Server User and Group Entries for
Synchronization
Add the ntUser and ntGroup object classes to any user and group entries, respectively, which will be
synchronized, along with any required attributes. Only Directory Server entries with those object classes
are synchronized. Active Directory entries which are synced over to Directory Server have those object
classes automatically.
Whenever the appropriate object classes are added to an entry, both for new entries and existing
entries, the entry is synced over at the next incremental update.
373
Red Hat D irectory Server 8.2 Administration Guide
Configuring Directory Server user entries for synchronization is described in Section 10.3.3, “Configuring
User Sync for Directory Server Users”, and configuring Directory Server group entries for
synchronization is described in Section 10.4.4, “Configuring Group Sync for Directory Server Groups”.
10.2.9. Step 9: Begin Synchronization
After the sync agreement is created, begin the synchronization process.
1. Go to the Configuration tab in the Console.
2. Open the Replication folder and expand the appropriate database.
3. Select the sync agreement.
4. Right-click on the agreement or open the Object menu.
5. Select Initiate Full Re-synchronization.
If synchronization stops for any reason, begin another total update (resynchronization) by selecting this
from the sync agreement menu. Beginning a total update (resynchronization) will not delete or overwrite
the databases.
10.3. Synchronizing Users
Users are not automatically synced between Directory Server and Active Directory. Synchronization both
directions has to be configured:
Users in the Active Directory domain are synced if it is configured in the sync agreement by selecting
the Sync New Windows Users option. All of the Windows users are copied to the Directory Server
when synchronization is initiated and then new users are synced over when they are created.
A Directory Server user account is synchronized to Active Directory through specific attributes that
are present on the Directory Server entry. Any Directory Server entry must have the ntUser object
class and the ntUserCreateNewAccount attribute; the ntUserCreateNewAccount attribute (even
on an existing entry) signals Directory Server Win Syn to write the entry over to the Active Directory
server.
New or modified user entries with the ntUser object class added are created and synced over to
the Windows machine at the next regular update, which is a standard poll of entry.
374
Chapter 10. Synchronizing Red Hat D irectory Server with Microsoft Active D irectory
NOTE
A user is not active on the Active Directory domain until it has a password. When an existing user
is modified to have the required Windows attributes, that user entry will be synced over to the
Active Directory domain, but will not be able to log in until the password is changed on the
Directory Server side or an administrator sets the password on Active Directory. T his is because
passwords stored in the Directory Server are encrypted, and Password Sync cannot sync already
encrypted passwords.
T o make the user active on the Active Directory domain, reset the user's password.
All synchronized entries in the Directory Server, whether they originated in the Directory Server or in
Active Directory, have special synchronization attributes:
ntUserDomainId. T his corresponds to the sAMAccountName attribute for Active Directory entries.
ntUniqueId. T his contains the value of the objectGUID attribute for the corresponding Windows
entry. T his attribute is set by the synchronization process and should not be set or modified
manually.
ntUserDeleteAccount. T his attribute is set automatically when a Windows entry is synced over but
must be set manually for Directory Server entries. If ntUserDeleteAccount has the value true, the
corresponding Windows entry be deleted when the Directory Server entry is deleted. Otherwise, the
entry remains in Active Directory, but is removed from the Directory Server database if it is deleted in
the Directory Server.
Setting ntUserCreateNewAccount and ntUserDeleteAccount on Directory Server entries allows the
Directory Manager precise control over which users within the synchronized subtree are synced on
Active Directory.
10.3.1. User Attributes Synchronized between Directory Server and Active
Directory
Only a subset of Directory Server and Active Directory attributes are synchronized. T hese attributes are
hard-coded and are defined regardless of which way the entry is being synchronized. Any other
attributes present in the entry, either in Directory Server or in Active Directory, remain unaffected by
synchronization.
Some attributes used in Directory Server and Active Directory are identical. T hese are usually attributes
defined in an LDAP standard, which are common among all LDAP services. T hese attributes are
synchronized to one another exactly. T able 10.3, “User Schema T hat Are the Same in Directory Server
and Windows Servers” shows attributes that are the same between the Directory Server and Windows
servers.
Some attributes define the same information, but the names of the attributes or their schema definitions
are different. T hese attributes are mapped between Active Directory and Directory Server, so that
attribute A in one server is treated as attribute B in the other. For synchronization, many of these
attributes relate to Windows-specific information. T able 10.2, “User Schema Mapped between Directory
Server and Active Directory” shows the attributes that are mapped between the Directory Server and
Windows servers.
For more information on the differences in ways that Directory Server and Active Directory handle some
schema elements, see Section 10.3.2, “User Schema Differences between Red Hat Directory Server and
Active Directory”.
375
Red Hat D irectory Server 8.2 Administration Guide
T able 10.2. User Schema Mapped between Directory Server and Active Directory
Directory Server
cn
[a]
Active Directory
name
ntUserDomainId
sAMAccountName
ntUserHomeDir
homeDirectory
ntUserScriptPath
scriptPath
ntUserLastLogon
lastLogon
ntUserLastLogoff
lastLogoff
ntUserAcctExpires
accountExpires
ntUserCodePage
codePage
ntUserLogonHours
logonHours
ntUserMaxStorage
maxStorage
ntUserProfile
profilePath
ntUserParms
userParameters
ntUserWorkstations
userWorkstations
[a] The cn is treated d ifferently than o ther s ync ed attrib utes . It is map p ed d irec tly (cn to cn) when s ync ing fro m Direc to ry Server
to Ac tive Direc to ry. When s ync ing fro m Ac tive Direc to ry to Direc to ry Server, ho wever, cn is map p ed fro m the name attrib ute o n
Wind o ws to the cn attrib ute in Direc to ry Server.
T able 10.3. User Schema T hat Are the Same in Directory Server and Windows Servers
cn
[a]
physicalDeliveryOfficeName
description
postOfficeBox
destinationIndicator
postalAddress
facsimileT elephoneNumber
postalCode
givenname
registeredAddress
homePhone
sn
homePostalAddress
st
initials
street
l
telephoneNumber
mail
teletexT erminalIdentifier
mobile
telexNumber
o
title
ou
usercertificate
pager
x121Address
[a] The cn is treated d ifferently than o ther s ync ed attrib utes . It is map p ed d irec tly (cn to cn) when s ync ing fro m Direc to ry Server
to Ac tive Direc to ry. When s ync ing fro m Ac tive Direc to ry to Direc to ry Server, ho wever, cn is map p ed fro m the name attrib ute o n
Wind o ws to the cn attrib ute in Direc to ry Server.
10.3.2. User Schema Differences between Red Hat Directory Server and Active
Directory
Although Active Directory supports the same basic X.500 object classes as Directory Server, there are a
few incompatibilities of which administrators should be aware.
10.3.2.1. Values for cn Attributes
376
Chapter 10. Synchronizing Red Hat D irectory Server with Microsoft Active D irectory
In Directory Server, the cn attribute can be multi-valued, while in Active Directory this attribute must have
only a single value. When the Directory Server cn attribute is synchronized, then, only one value is sent
to the Active Directory peer.
What this means for synchronization is that,potentially, if a cn value is added to an Active Directory entry
and that value is not one of the values for cn in Directory Server, then all of the Directory Server cn
values are overwritten with the single Active Directory value.
One other important difference is that Active Directory uses the cn attribute attribute as its naming
attribute, where Directory Server uses uid. T his means that there is the potential to rename the entry
entirely (and accidentally) if the cn attribute is edited in the Directory Server. If that cn change is written
over to the Active Directory entry, then the entry is renamed, and the new named entry is written back
over to Directory Server.
10.3.2.2. Password Policies
Both Active Directory and Directory Server can enforce password policies such as password minimum
length or maximum age. Windows Sync makes no attempt to ensure that the policies are consistent,
enforced, or synchronized. If password policy is not consistent in both Directory Server and Active
Directory, then password changes made on one system may fail when synced to the other system. T he
default password syntax setting on Directory Server mimics the default password complexity rules that
Active Directory enforces.
10.3.2.3. Values for street and streetAddress
Active Directory uses the attribute streetAddress for a user or group's postal address; this is the way
that Directory Server uses the street attribute. T here are two important differences in the way that
Active Directory and Directory Server use the streetAddress and street attributes, respectively:
In Directory Server, streetAddress is an alias for street. Active Directory also has the street
attribute, but it is a separate attribute that can hold an independent value, not an alias for
streetAddress.
Active Directory defines both streetAddress and street as single-valued attributes, while
Directory Server defines street as a multi-valued attribute, as specified in RFC 4519.
Because of the different ways that Directory Server and Active Directory handle streetAddress and
street attributes, there are two rules to follow when setting address attributes in Active Directory and
Directory Server:
Windows Sync maps streetAddress in the Windows entry to street in Directory Server. T o avoid
conflicts, the street attribute should not be used in Active Directory.
Only one Directory Server street attribute value is synced to Active Directory. If the streetAddress
attribute is changed in Active Directory and the new value does not already exist in Directory Server,
then all street attribute values in Directory Server are replaced with the new, single Active Directory
value.
10.3.2.4 . Constraints on the initials Attribute
For the initials attribute, Active Directory imposes a maximum length constraint of six characters, but
Directory Server does not have a length limit. If an initials attribute longer than six characters is
added to Directory Server, the value is trimmed when it is synchronized with the Active Directory entry.
10.3.3. Configuring User Sync for Directory Server Users
For Directory Server users to be synchronized over to Active Directory, the user entries must have the
377
Red Hat D irectory Server 8.2 Administration Guide
appropriate sync attributes set.
10.3.3.1. Configuring User Sync in the Console
1. In the Directory Server Console, select the Directory tab.
2. For an existing entry, right-click the entry, and click Properties to open the property editor for
the entry.
For a new entry, right-click the main entry in the left window to add the new entry, select User, and
then fill in the required entry attributes.
3. On the left side of the Property Editor, click the NT User link.
4. In the NT User tab, check the Enable NT Attributes checkbox.
5. T o enable synchronization, two fields are required:
Setting a NT User ID
Selecting the Create New NT Account checkbox
6. Selecting the Delete NT Account checkbox means that the corresponding Windows user is
deleted if the Directory Server entry is deleted.
7. Set the other Windows attributes. T hese attributes are mapped to relevant Windows attributes.
Additional ntUser attributes can be created either by using the Advanced button; see
Section 3.2.4.2, “Modifying Entries Using ldapmodify”.
378
Chapter 10. Synchronizing Red Hat D irectory Server with Microsoft Active D irectory
TIP
Reset the user's password.
A user is not active on the Active Directory domain until it has a password. When an existing user
is modified to have the required Windows attributes, that user entry will be synced over to the
Active Directory domain, but will not be able to log in until the password is changed on the
Directory Server side or an administrator sets the password on Active Directory. Password Sync
cannot sync encrypted passwords.
So, to make the user active on the Active Directory domain, reset the user's password.
10.3.3.2. Configuring User Sync in the Command Line
T o enable synchronization through the command line, add the required sync attributes to an entry or
create an entry with those attributes.
T hree schema elements are required for synchronization:
T he ntUser object class
T he ntUserDomainId attribute, to give the Windows ID
T he ntUserCreateNewAccount attribute, to signal to the synchronization plug-in to sync the
Directory Server entry over to Active Directory
For example:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: uid=scarter,ou=People,dc=example,dc=com
changetype: modify
add: objectClass
objectClass:ntUser
add: ntUserDomainId
ntUserDomainId: Sam Carter
add: ntUserCreateNewAccount
ntUserCreateNewAccount: true
add: ntUserDeleteAccount
ntUserDeleteAccount: true
Many additional Windows and user attributes can be added to the entry. All of the schema which is
synchronized is listed in Section 10.3.1, “User Attributes Synchronized between Directory Server and
Active Directory”. Windows-specific attributes, belonging to the ntUser object class, are described in
more detail in the Schema Reference.
379
Red Hat D irectory Server 8.2 Administration Guide
TIP
Reset the user's password.
A user is not active on the Active Directory domain until it has a password. When an existing user
is modified to have the required Windows attributes, that user entry will be synced over to the
Active Directory domain, but will not be able to log in until the password is changed on the
Directory Server side or an administrator sets the password on Active Directory. Password Sync
cannot sync encrypted passwords.
So, to make the user active on the Active Directory domain, reset the user's password.
10.3.4. Configuring User Sync for Active Directory Users
Synchronization for Windows users (users which originate in the Active Directory domain) is configured
in the sync agreement.
10.3.4 .1. Configuring User Sync in the Console
1. Open the Configuration tab and expand the Replication folder.
2. Open the appropriate database, and select the sync agreement.
3. Open the Connection tab.
4. Check the New Windows User Sync checkbox to enable users sync. T o disable sync, uncheck
the box.
For new sync agreements, select the corresponding users sync checkbox in the sync agreement
creation wizard.
10.3.4 .2. Configuring User Sync in the Command Line
T he attribute to set Active Directory user sync is nsds7NewWinUserSyncEnabled and is set on the sync
agreement. T o enable user sync, add this attribute to the sync agreement or create a sync agreement
with this attribute set to on:
380
Chapter 10. Synchronizing Red Hat D irectory Server with Microsoft Active D irectory
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=ExampleSyncAgreement,cn=userRoot,cn=dc=example\,dc=com,cn=mapping
tree,cn=config
changetype: modify
replace: nsds7NewWinUserSyncEnabled
nsds7NewWinUserSyncEnabled: on
T o disable user sync, set nsds7NewWinUserSyncEnabled: off.
10.4. Synchronizing Groups
Like user entries, groups are not automatically synced between Directory Server and Active Directory.
Synchronization both directions has to be configured:
Groups in the Active Directory domain are synced if it is configured in the sync agreement by
selecting the Sync New Windows Groups option. All of the Windows groups are copied to the
Directory Server when synchronization is initiated and then new groups are synced over as they are
created.
A Directory Server group account is synchronized to Active Directory through specific attributes that
are present on the Directory Server entry. Any Directory Server entry must have the ntGroup object
class and the ntGroupCreateNewGroup attribute; the ntGroupCreateNewGroup attribute (even on
an existing entry) signals Directory Server Win Syn to write the entry over to the Active Directory
server.
New or modified groups that have the ntGroup object class are created and synced over to the
Windows machine at the next regular update.
IMPORTANT
When a group is synced, the list of all of its members is also synced. However, the member
entries themselves are not synced unless user sync is enabled and applies to those entries.
T his could created a problem is an application or service tries to do a modify operation on all
members in a group on the Active Directory server, and some of those users do not exist.
Additionally, groups have a few other common attributes:
T wo attributes control whether Directory Server groups are created and deleted on Active Directory,
ntGroupCreateNewGroup and ntGroupDeleteGroup.
ntGroupCreateNewGroup is required to sync Directory Server groups over to Active Directory.
ntUserDomainId contains the unique ID for the entry on the Active Directory domain. T his is the only
required attribute for the ntGroup object class.
ntGroupType is the type of Windows group. Windows group types are global/security, domain
local/security, global/distribution, or domain local/distribution. T his is set automatically for Windows
groups that are synchronized over, but this attribute must be set manually on Directory Server
entries before they can be synced.
10.4.1. About Windows Group Types
In Active Directory, there are two major types of groups: security and distribution. Security groups are
most similar to groups in Directory Server, since security groups can have policies configured for access
381
Red Hat D irectory Server 8.2 Administration Guide
controls, resource restrictions, and other permissions. Distribution groups are for mailing distribution.
T hese are further broken down into global and local groups. T he Directory Server ntGroupType
supports all four group types:
-214 8364 6 for global/security (the default)
-214 8364 4 for domain local/security
2 for global/distribution
4 for domain local/distribution
10.4.2. Group Attributes Synchronized between Directory Server and Active
Directory
Only a subset of Directory Server and Active Directory attributes are synchronized. T hese attributes are
hard-coded and are defined regardless of which way the entry is being synchronized. Any other
attributes present in the entry, either in Directory Server or in Active Directory, remain unaffected by
synchronization.
Some attributes used in Directory Server and Active Directory group entries are identical. T hese are
usually attributes defined in an LDAP standard, which are common among all LDAP services. T hese
attributes are synchronized to one another exactly. T able 10.5, “Group Entry Attributes T hat Are the
Same between Directory Server and Active Directory” shows attributes that are the same between the
Directory Server and Windows servers.
Some attributes define the same information, but the names of the attributes or their schema definitions
are different. T hese attributes are mapped between Active Directory and Directory Server, so that
attribute A in one server is treated as attribute B in the other. For synchronization, many of these
attributes relate to Windows-specific information. T able 10.4, “Group Entry Attribute Mapping between
Directory Server and Active Directory” shows the attributes that are mapped between the Directory
Server and Windows servers.
For more information on the differences in ways that Directory Server and Active Directory handle some
schema elements, see Section 10.4.3, “Group Schema Differences between Red Hat Directory Server
and Active Directory”.
T able 10.4 . Group Entry Attribute Mapping between Directory Server and Active Directory
Directory Server
Active Directory
cn
name
ntUserDomainID
name
ntGroupT ype
groupT ype
uniqueMember
Member
[a]
member
[a] The Member attrib ute in Ac tive Direc to ry is s ync ed to the uniqueMember attrib ute in Direc to ry Server.
382
Chapter 10. Synchronizing Red Hat D irectory Server with Microsoft Active D irectory
T able 10.5. Group Entry Attributes T hat Are the Same between Directory Server and Active
Directory
cn
o
description
ou
l
seeAlso
mail
10.4.3. Group Schema Differences between Red Hat Directory Server and Active
Directory
Although Active Directory supports the same basic X.500 object classes as Directory Server, there are a
few incompatibilities of which administrators should be aware.
Nested groups (where a group contains another group as a member) are supported and for WinSync
are synchronized. However, Active Directory imposes certain constraints as to the composition of nested
groups. For example, a global group contain a domain local group as a member. Directory Server has no
concept of local and global groups, and, therefore, it is possible to create entries on the Directory Server
side that violate Active Directory's constraints when synchronized.
10.4.4. Configuring Group Sync for Directory Server Groups
For Directory Server groups to be synchronized over to Active Directory, the group entries must have the
appropriate sync attributes set.
10.4 .4 .1. Configuring Group Sync in the Console
1. In the Directory Server Console, select the Directory tab.
2. Right-click the group entry, and click Advanced to open the advanced property editor for the
entry. All of the sync-related attributes must be added manually, so only the advanced property
editor can set the attributes.
3. Click the objectClasses field, and then click the Add Value button.
4. Select the ntGroup object class.
383
Red Hat D irectory Server 8.2 Administration Guide
5. Setting the ntGroup object class automatically adds the ntUserDomainId attribute. T his attribute
is required, so add a value.
6. T o enable synchronization, click the Add Attribute button, and select the
ntGroupCreateNewGroup attribute from the list. T hen, set its value to on. T his signals to the
sync plug-in that the entry should be added to the Active Directory directory.
T o delete the group entry from the Active Directory domain if it is deleted from the Directory Server
database, set the ntGroupDeleteGroup attribute and set it to on.
7. Add any other Windows attributes for the Directory Server entry. T he available attributes are listed
in Section 10.4.2, “Group Attributes Synchronized between Directory Server and Active Directory”.
If the ntGroupType is not added, then the group is automatically added as a global security group
(ntGroupT ype:-214 8364 6).
10.4 .4 .2. Configuring Group Sync in the Command Line
T o enable synchronization through the command line, add the required sync attributes to an entry or
384
Chapter 10. Synchronizing Red Hat D irectory Server with Microsoft Active D irectory
create an entry with those attributes.
T hree schema elements are required for synchronization:
T he ntGroup object class.
T he ntUserDomainId attribute, to give the Windows ID for the entry.
T he ntGroupCreateNewGroup attribute, to signal to the synchronization plug-in to sync the Directory
Server entry over to Active Directory.
T he ntGroupDeleteGroup attribute is optional, but this sets whether to delete the entry
automatically from the Active Directory domain if it is deleted in the Directory Server.
It is also recommended to the ntGroupType attribute. If this attribute is not specified, then the group is
automatically added as a global security group (ntGroupT ype:-214 8364 6).
For example:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=Example Group,ou=Groups,dc=example,dc=com
changetype: modify
add: objectClass
objectClass:ntGroup
add: ntUserDomainId
ntUserDomainId: example-group
add: ntGroupCreateNewGroup
ntGroupCreateNewGroup: true
add: ntGroupDeleteGroup
ntGroupDeleteGroup: true
add: ntGroupType
ntGroupType: 2
Many additional Windows and group attributes can be added to the entry. All of the schema which is
synchronized is listed in Section 10.4.2, “Group Attributes Synchronized between Directory Server and
Active Directory”. Windows-specific attributes, belonging to the ntGroup object class, are described in
more detail in the Schema Reference.
10.4.5. Configuring Group Sync for Active Directory Groups
Synchronization for Windows users (users which originate in the Active Directory domain) is configured
in the sync agreement.
10.4 .5.1. Configuring Group Sync in the Console
1. Open the Configuration tab and expand the Replication folder.
2. Open the appropriate database, and select the sync agreement.
385
Red Hat D irectory Server 8.2 Administration Guide
3. Open the Connection tab.
4. Check the New Windows Group Sync checkbox to enable group sync. T o disable sync,
uncheck the box.
For new sync agreements, select the corresponding group sync checkbox in the sync agreement
creation wizard.
10.4 .5.2. Configuring Group Sync in the Command Line
T he attribute to set Active Directory group sync is nsds7NewWinGroupSyncEnabled and is set on the
sync agreement. T o enable group sync, add this attribute to the sync agreement or create a sync
agreement with this attribute set to on:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=ExampleSyncAgreement,cn=userRoot,cn=dc=example\,dc=com,cn=mapping
tree,cn=config
changetype: modify
replace: nsds7NewWinGroupSyncEnabled
nsds7NewWinGroupSyncEnabled: on
T o disable group sync, set nsds7NewWinGroupSyncEnabled: off.
10.5. Deleting and Resurrecting Entries
T his section describes how enabling synchronization affects deleted entries on the sync peers and how
resurrected entries are handled.
10.5.1. Deleting Entries
All changes on an Active Directory peers are always synced back to the Directory Server. T his means
that when an Active Directory group or user account is deleted on the Active Directory domain, the
deletion is automatically synced back to the Directory Server sync peer server.
On Directory Server, on the other hand, when a Directory Server account is deleted, the corresponding
386
Chapter 10. Synchronizing Red Hat D irectory Server with Microsoft Active D irectory
entry on Active Directory is only deleted if the Directory Server entry has the ntUserDeleteAccount or
ntGroupDeleteGroup attribute set to true.
NOTE
When a Directory Server entry is synchronized over to Active Directory for the first time, Active
Directory automatically assigns it a unique ID. At the next synchronization interval, the unique ID
is synchronized back to the Directory Server entry and stored as the ntUniqueId attribute. If the
Directory Server entry is deleted on Active Directory before the unique ID is synchronized back to
Directory Server, the entry will not be deleted on Directory Server. Directory Server uses the
ntUniqueId attribute to identify and synchronize changes made on Active Directory to the
corresponding Directory Server entry; without that attribute, Directory Server will not recognize the
deletion.
T o delete the entry on Active Directory and then synchronize the deletion over to Directory
Server, wait the length of the winSyncInterval (by default, five minutes) after the entry is
created before deleting it so that the ntUniqueId attribute is synchronized.
10.5.2. Resurrecting Entries
It is possible to add deleted entries back in Directory Server; the deleted entries are called tombstone
entries. When a deleted entry which was synced between Directory Server and Active Directory is readded to Directory Server, the resurrected Directory Server entry has all of its original attributes and
values. T his is called tombstone reanimation. T he resurrected entry includes the original ntUniqueId
attribute which was used to synchronize the entries, which signals to the Active Directory server that this
new entry is a tombstone entry.
T he way that tombstone entries are handled is different between Windows Server 2000 and Windows
Server 2003:
On Windows 2000, Active Directory creates a new entry with a new unique ID; this new ID is synced
back to the Directory Server entry.
On Windows 2003, Active Directory resurrects the old entry and preserves the original unique ID for
the entry.
For Active Directory entries on both on Windows 2000 and 2003, when the tombstone entry is
resurrected on Directory Server, all of the attributes of the original Directory Server are retained and are
still included in the resurrected Active Directory entry.
10.6. Sending Synchronization Updates
Synchronization occurs as frequently as is set in the winSyncInterval setting (for retrieving changes
from the Active Directory domain) or nsds5replicaupdateschedule setting (for pushing changes from
the Directory Server). By default, changes are retrieved from Active Directory every five minutes, and
changes from the Directory Server are sent immediately.
A sync update can be triggered manually. It is also possible to do a full resynchronization, which sends
and pulls every entry in the Directory Server and Active Directory as if it were new. A full
resynchronization includes existing Directory Server entries which may not have previously been
synchronized.
10.6.1. Performing a Manual Sync Update
During normal operations, all the updates made to entries in the Directory Server that need to be sent to
387
Red Hat D irectory Server 8.2 Administration Guide
During normal operations, all the updates made to entries in the Directory Server that need to be sent to
Active Directory are collected the changelog and then replayed during an incremental update.
1. Go to the Configuration tab in the Console.
2. Open the Replication folder and expand the appropriate database.
3. Select the sync agreement.
4. Right-click on the agreement or open the Object menu.
5. Select Send and Receive Updates from the drop down menu.
10.6.2. Sending a Total Update (Full Synchronization)
If there have been major changes to data, or synchronization attributes are added to pre-existing
Directory Server entries, it is necessary to initiate a resynchronization. Resynchronization is a total
update; the entire contents of synchronized subtrees are examined and, if necessary, updated.
Resynchronization is done without using the changelog. T his is similar to initializing or reinitializing a
consumer in replication.
1. Go to the Configuration tab in the Console.
2. Open the Replication folder and expand the appropriate database.
3. Select the sync agreement.
4. Right-click on the agreement or open the Object menu.
5. Select Initialize Full Re-synchronization from the drop down menu.
Resynchronizing will not delete data on the sync peer; it sends and receives all updates and add
any new or modified Directory Server entries; for example, it adds a pre-existing Directory Server
user that had the ntUser object class added.
388
Chapter 10. Synchronizing Red Hat D irectory Server with Microsoft Active D irectory
10.6.3. Sending Sync Updates in the Command Line
T o send sync updates through the command line, add the nsDS5BeginReplicaRefresh attribute to the
sync agreement. For example:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=ExampleSyncAgreement,cn=sync replica,cn=dc=example\,dc=com,cn=mapping
tree,cn=config
changetype: modify
add: nsDS5BeginReplicaRefresh
nsDS5BeginReplicaRefresh: start
T his attribute is removed from the entry as soon as the update is complete.
NOTE
T his initiates a full synchronization for the entire database, not an incremental update of recent
changes.
10.6.4. Checking Synchronization Status
Check synchronization status in the Replication tab in the Status of the Console. Highlight the
synchronization agreement to monitor, and the relevant information should appear in the right-hand
pane. T he Status area shows whether the last incremental and total updates were successful and
when they occurred.
1. Go to the Configuration tab in the Console.
2. Open the Replication folder and expand the appropriate database.
3. Select the sync agreement.
4. In the Sum m ary tab, the status of the latest sync process is shown at the bottom.
389
Red Hat D irectory Server 8.2 Administration Guide
10.7. Modifying the Sync Agreement
Certain attributes of the sync agreement can be modified, including the connection information. Using the
command line, many additional parameters can be created with or added to the sync agreement,
including changing the sync interval and setting a sync schedule.
10.7.1. Editing the Sync Agreement in the Console
Most of the information which can be edited in the Console is limited to connection information, including
the protocol to use and the bind credentials. It is also possible to edit the sync agreement description.
1. In the Configuration tab, expand the Replication folder.
2. Expand the database being synced. All of the synchronization agreements are listed below the
database. Double-click the sync agreement to open it in the main window.
3. Click the Connection tab.
390
Chapter 10. Synchronizing Red Hat D irectory Server with Microsoft Active D irectory
T here are three areas of information that can be edited.
T he connection type (standard, T LS/SSL, and Start T LS).
T he bind user, both DN and password.
Whether to sync new Directory Server users and new Directory Server groups automatically.
T here are three options for the connection type — standard, T LS/SSL, and Start T LS — but there
are really only two connection protocols, LDAP and LDAPS. Both a standard connection and Start
T LS connection use LDAP (Start T LS creates a secure connection over an insecure port).
It is not possible to change the connection protocol because it is not possible to change the port
number used to connect to the Windows sync peer.
It is possible to change the connection type between the standard connection and Start T LS, but it
is not possible to change from T LS/SSL to either the standard or Start T LS connections. Likewise,
it is not possible to go from standard or Start T LS to T LS/SSL. If you need to change the
connection protocol or the port number, delete the sync agreement and create a new one.
10.7.2. Adding and Editing the Sync Agreement in the Command Line
Creating or editing the sync agreement through the command line is more flexible and provides more
options than using the Directory Server Console. T he full list of sync agreement attributes are listed in
Section 10.7.2.4, “Sync Agreement Attributes” and described in the Configuration and Command-Line
Tool Reference.
10.7.2.1. Creating a Basic Sync Agreement
T he most basic sync agreement defines the Directory Server database and the Active Directory sync
peer:
For the Directory Server database:
T he replicated subtree in the directory (nsds7DirectoryReplicaSubtree)
T he Directory Server root DN (nsDS5ReplicaRoot)
T he replicated subtree in the directory (nsds7DirectoryReplicaSubtree)
For the Active Directory domain:
T he replicated subtree in the Active Directory domain (nsds7WindowsReplicaSubtree)
T he Active Directory domain name (nsds7WindowsDomain)
It also defines the connection information that the Directory Server uses to bind to the Active Directory
391
Red Hat D irectory Server 8.2 Administration Guide
domain:
T he Active Directory hostname (nsDS5ReplicaHost).
T he Active Directory port (nsDS5ReplicaPort).
T he type of connection (nsDS5ReplicaTransportInfo), which can be standard (LDAP), SSL (SSL),
or Start T LS (T LS), which is a secure connection over a standard port.
T he username (nsDS5ReplicaBindDN) and password (nsDS5ReplicaBindCredentials) for the
Directory Server to use to bind to the Active Directory server.
For example:
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=ExampleSyncAgreement,cn=sync replica,cn=dc=example\,dc=com,cn=mapping
tree,cn=config
changetype: add
objectclass: top
objectclass: nsDSWindowsReplicationAgreement
cn: ExampleSyncAgreement
nsds7WindowsReplicaSubtree: cn=Users,dc=ad1
nsds7DirectoryReplicaSubtree: ou=People,dc=example,dc=com
nsds7WindowsDomain: ad1
nsDS5ReplicaRoot: dc=example,dc=com
nsDS5ReplicaHost: ad1.windows-server.com
nsDS5ReplicaPort: 389
nsDS5ReplicaBindDN: cn=sync user,cn=config
nsDS5ReplicaBindCredentials: {DES}ffGad646dT0nnsT8nJOaMA==
nsDS5ReplicaTransportInfo: TLS
nsds7NewWinUserSyncEnabled: on
nsds7NewWinGroupSyncEnabled: on
10.7.2.2. Setting Sync Schedules
Synchronization works two ways. T he Directory Server sends its updates to Active Directory on a
configurable schedule, similar to replication, using the nsds5replicaupdateschedule attribute. T he
Directory Server polls the Active Directory to check for changes; the frequency that it checks the Active
Directory server is set in the winSyncInterval attribute.
By default, the Directory Server update schedule is to always be in sync. T he Active Directory interval is
to poll the Active Directory every five minutes.
T o change the schedule the Directory Server uses to send its updates to the Active Directory, edit the
nsds5replicaupdateschedule attribute. T he schedule is set with start (SSSS) and end (EEEE) times
in the form HHMM, using a 24-hour clock. T he days to schedule sync updates are use ranging from 0
(Sunday) to 6 (Saturday).
nsds5replicaupdateschedule: SSSS EEEE DDDDDDD
For example, this schedules synchronization to run from noon to 2:00pm on Sunday, T uesday,
T hursday, and Saturday:
nsds5replicaupdateschedule: 1200 1400 0246
392
Chapter 10. Synchronizing Red Hat D irectory Server with Microsoft Active D irectory
NOTE
T he synchronization times cannot wrap around midnight, so the setting 2300 0100 is not valid.
T o change how frequently the Directory Server checks the Active Directory for changes to Active
Directory entries, reset the winSyncInterval attribute. T his attribute is set in seconds, so the default of
300 means that the Directory Server polls the Active Directory server every 300 seconds, or five
minutes. Setting this to a higher value can be useful if the directory searches are taking too long and
affecting performance.
winSyncInterval: 1000
10.7.2.3. Changing Sync Connections
T wo aspects of the connection for the sync agreement can be altered:
T he bind username and password (nsDS5ReplicaBindDN and nsDS5ReplicaBindCredentials).
T he connection method (nsDS5ReplicaTransportInfo).
It is only possible to change the nsDS5ReplicaTransportInfo from LDAP to T LS and vice versa. It
is not possible to change to or from SSL because it is not possible to change the port number, and
switching between LDAP and LDAPS requires changing the port number.
For example:
nsDS5ReplicaBindDN: cn=sync user,cn=config
nsDS5ReplicaBindCredentials: {DES}ffGad646dT0nnsT8nJOaMA==
nsDS5ReplicaTransportInfo: TLS
CAUTION
It is not possible to change the port number of the Active Directory sync peer. T herefore, it is also
not possible to switch between standard/Start T LS connections and SSL connections, since that
requires changing between standard and insecure ports.
T o change to or from SSL/T LS, delete the sync agreement and add it again with the updated port
number and new transport information.
10.7.2.4 . Sync Agreement Attributes
T he common sync agreement attributes are listed in T able 10.6, “Sync Agreement Attributes”. All of the
possible sync agreement attributes are described in detail in the Configuration and Command-Line Tool
Reference and Schema Reference.
393
Red Hat D irectory Server 8.2 Administration Guide
T able 10.6. Sync Agreement Attributes
Object Class or Attribute
Description
nsDSWindowsReplicationAgreement
An operational object class which contains the
sync agreement attributes.
cn
Gives the name for the sync agreement.
nsds7WindowsReplicaSubtree
Specifies the Windows server suffix (root or sub)
that is synced.
nsds7DirectoryReplicaSubtree
Specifies the Directory Server suffix (root or sub)
that is synced.
nsds7NewWinUserSyncEnabled
Sets whether new Windows user accounts are
automatically created on the Directory Server.
nsds7NewWinGroupSyncEnabled
Specifies whether new Windows group accounts
are automatically created on the Directory Server.
nsds7WindowsDomain
Identifies the Windows domain being
synchronized; analogous to nsDS5ReplicaHost
in a replication agreement.
nsds5replicaport
Gives the LDAP port for the Windows server.
T o use T LS/SSL, give the secure port number
(636 by default) and set the
nsds5ReplicaTransportInfo attribute to SSL.
T o use Start T LS, which initiates a secure
connection over a standard port, use the
standard port, 389, with the
nsds5ReplicaTransportInfo attribute to T LS.
nsds5replicatransportinfo
T o use T LS/SSL, set this parameter to SSL.
T o use Start T LS, which initiates a secure
connection over a standard port, set this
parameter to T LS.
T o use simple authentication, set this parameter
to LDAP.
nsds5ReplicaBindDN
T he sync user DN used by the Directory Server
instance to bind to the Windows server.
nsds5replicabindmethod
T he connection type for replication between the
servers. T he connection type defines how the
supplier authenticates to the consumer.
Leaving the bind method empty or setting it to
SIMPLE means that the server uses basic
password-based authentication. T his requires
the nsds5ReplicaBindDN and
nsds5ReplicaCredentials attributes to give
the bind information.
T he SSLCLIENT AUT H option uses a secure
connection. T his requires setting the
nsds5ReplicaTransportInfo attribute be set to
394
Chapter 10. Synchronizing Red Hat D irectory Server with Microsoft Active D irectory
SSL or T LS.
nsds5replicabindcredentials
Only for simple authentication. Stores the hashed
password used with the bind DN given for simple
authentication.
nsds5replicaroot
Sets which Directory Server subtree is replicated.
Usually, it is recommended that the replicated
subtree be high in the directory tree so that the
entire database is replicated. For example:
dc=example,dc=com
description
A text description of the replication agreement.
Make this a useful description so it is easier to
manage sync agreements.
nsds5replicaupdateschedule
Sets the start and end time for the replication
updates and the days on which replication occurs
in the form start_time end_time days. If the
schedule is omitted, synchronization occurs all of
the time.
winSyncInterval
Optional. Sets how frequently, in seconds, the
Directory Server polls the Windows server for
updates to write over. If this is not set, the default
is 300, which is 300 seconds or five (5) minutes.
nsds5BeginReplicaRefresh
Optional. Performs an online (immediate)
initialization of the sync peer. If this is set, the
attribute is only present as long as the sync peer
is being updated initialized; when the initialization
is complete, the attribute is deleted automatically.
T he only value when adding this attribute is
start.
10.8. Configuring Unidirectional Synchronization
Synchronization has to be configured both ways. T his means it is also possible to configure
synchronization only one way.
Directory Server synchronization is set by adding the appropriate attributes on the individual directory
entries. T o configure synchronization from the Directory Server database to the Active Directory server,
set the appropriate ntUser and ntGroup object classes and attributes on the user and group entries,
respectively.
Active Directory synchronization is configured in the sync agreement. T o configure synchronization from
Active Directory to the Directory Server database, then set the New Windows Users Sync and New
Windows Groups Sync attributes in the synchronization agreement.
395
Red Hat D irectory Server 8.2 Administration Guide
NOTE
A synchronization agreement needs to be configured for both kinds of unidirectional
synchronization.
T o only sync Directory Server entries, then do not set the Active Directory sync attributes in the sync
agreement. Likewise, to only synchronize Active Directory over to Directory Server, do not add any sync
attributes to Directory Server entries.
10.9. Managing the Password Sync Service
T he Password Sync Service must be installed on every Active Directory domain controller. T he service
synchronizes password changes made on Active Directory with the corresponding entries' passwords
on the Directory Server. Like any Windows service, it can be modified, started and stopped, and
uninstalled, depending on how synchronization between Directory Server and Active Directory changes.
10.9.1. Modifying Password Sync
T o reconfigure Password Sync:
1. Open Control Panel, and double-click Add/Rem ove Program s.
2. Click the Change button to relaunch the installer to change the settings.
3. Go back through the configuration screens to make any changes to the configuration.
10.9.2. Starting and Stopping the Password Sync Service
T he Password Sync Service is configured to start whenever the Active Directory host is started. T o
reconfigure the service so that it does not start when Windows reboots:
1. Go to the Control Panel, and select Services.
2. Scroll through the list of services for the Password Sync Service. T he Startup field is set to
Autom atic.
3. Double-click on Password Sync.
4. Select the Manual radio button, and then click OK.
396
Chapter 10. Synchronizing Red Hat D irectory Server with Microsoft Active D irectory
T o start and stop Password Sync:
1. Go to the Control Panel, and select Services.
2. Scroll through the list of services for Password Sync, and right-click.
3. Select Stop, Start, or Restart, and hit okay.
397
Red Hat D irectory Server 8.2 Administration Guide
It's also possible to select the sync service and then click the start or stop links in the upper left of
the Services window.
Changed passwords are captured even if Password Sync is not running. If Password Sync is restarted,
the password changes are sent to Directory Server at the next synchronization.
10.9.3. Uninstalling Password Sync Service
1. Open Control Panel, and double-click Add/Rem ove Program s.
2. Select click Rem ove to uninstall the Password Sync Service.
3. If SSL was configured for the Password Sync, then the cert8.db and key3.db databases that
were created were not removed when Password Sync was uninstalled. Delete these files by hand.
398
Chapter 10. Synchronizing Red Hat D irectory Server with Microsoft Active D irectory
10.9.4. Upgrading Password Sync
T he Password Sync service cannot be upgraded directly. However, the existing certificates, keys, and
configuration can be applied to the new service if the new service is installed before the old one is
removed. T hen, it is not necessary to reconfigure the service like new; it picks up the information it
needs from the registry.
1. Download the PassSync.m si file from the appropriate Directory Server channel in Red Hat
Network and save it to the Active Directory machine.
NOTE
T here are two PassSync packages available, one for 32-bit Windows servers and one for
64-bit. Make sure to select the appropriate packages for your Windows platform.
2. Double-click on the PassSync.m si file to install it.
3. All of the previous information should be included, so click Finish to install the new Password
Sync.
T he previous SSL certificates and configuration is also preserved, so it is not necessary to
reconfigure SSL.
4. Open the Add/Rem ove Program s window.
5. Select the older version of Password Sync and click the Rem ove button.
NOTE
Check the version numbers to make sure the right Password Sync service is removed.
6. Reboot the Windows machine to start Password Sync.
NOTE
T he Windows machine must be rebooted. Without the rebooting, PasswordHook.dll is
not enabled, and password synchronization will not function.
10.10. Troubleshooting
If synchronization does not seem to be functioning properly, see the Windows event log and/or Directory
Server error log for information on any potential problems.
Enable replication logging for more detailed information on synchronization to be recorded in the error
logs. Replication log levels produces more verbose logs from the sync code that can help in diagnosing
problems.
1. In the Console, click the Configuration tab, select Logs from the navigation menu on the right,
and open the error log.
2. Scroll down to error log level, and select Replication from the menu. Hit save.
For complete information on error log levels, refer to Red Hat Directory Server Configuration and
Command-Line Tool Reference.
399
Red Hat D irectory Server 8.2 Administration Guide
Error #1: T he message box when creating the sync agreement indicates that the it cannot
connect to Active Directory.
Make sure that the directory suffixes, Windows domain and domain host, and the administrator DN and
password are correct. Also verify that the port numbers used for LDAPS is correct. If all of this is correct,
make sure that Active Directory or the Windows machine are running.
Error #2: After synchronization, the status returns error 81.
One of the sync peer servers has not been properly configured for SSL communication. Examine the
Directory Server access log file to see if the connection attempt was received by the Directory Server.
T here are also helpful messages in the Directory Server's error log file.
T o narrow down the source of the misconfiguration, try to establish an LDAPS connection to the
Directory Server. If this connection attempt fails, check all values (port number, hostname, search base,
and so forth) to see if any of these are the problem. If all else fails, reconfigure the Directory Server with
a new certificate.
If the LDAPS connection is successful, it is likely that the misconfiguration is on Active Directory. Examine
the Windows event log file for error messages.
NOTE
A common problem is that the certificate authority was not configured as trusted when the
Windows sync services certificate database was configured.
Error #3: An entry is moved from one subtree on Active Directory to another subtree, but the
user is not moved to the corresponding subtree on Directory Server.
T his is a known issue with synchronizing modrdn operations on Active Directory with entries on
Directory Server. T o work around it, delete the entry on Active Directory and then add it anew to the new
subtree. T he deletion and the addition will be properly synchronized to the Directory Server peer.
4 00
Chapter 11. Organizing Entries with Groups, Roles, and Views
Chapter 11. Organizing Entries with Groups, Roles, and Views
Entries contained within the directory can be grouped in different ways to simplify the management of
user accounts. Red Hat Directory Server supports a variety of methods for grouping entries and sharing
attributes between entries. T o take full advantage of the features offered by roles and class of service,
determine the directory topology when planning the directory deployment.
11.1. Using Groups
Groups are a mechanism for associating entries for ease of administration. Groups do not have the
flexibility or utility of roles. However, there are certain clients and applications where groups are useful.
Groups also offer compatibility with older LDAP clients and directory services.
NOTE
Managing groups is significantly easier by using the memberOf attribute to identify in user entries
to what groups a user belongs. T he memberOf attribute is maintained by the Directory Server and
updated automatically on entries as group membership changes. See Section 11.1.4, “Using the
memberOf Attribute to Manage Group Membership Information” for information on using the
memberOf attribute.
11.1.1. Creating Static Groups in the Console
Static groups organize entries by specifying the same group value in the DN attribute of any number of
users.
NOTE
If a user has an entry on a remote Directory Server (for example, in a chained database), different
from the Directory Server which has the entry that defines the static group, then use the
Referential Integrity plug-in to ensure that deleted user entries are automatically deleted from the
static group.
T here are some performance and access control considerations with the Referential Integrity
plug-in. For more information about using referential integrity with chaining, see Section 2.3.2,
“Configuring the Chaining Policy”.
1. In the Directory Server Console, select the Directory tab.
2. In the left pane, right-click the entry under which to add a new group, and select New > Group.
4 01
Red Hat D irectory Server 8.2 Administration Guide
Alternatively, go to the Object menu, and select New > Group.
3. Click General in the left pane. T ype a name for the new group in the Group Nam e field (the
name is required), and enter a description of the new group in the Description field.
4. Click Mem bers in the left pane. In the right pane, select the Static Group tab. Click Add to add
new members to the group.
5. In the Search drop-down list, select what sort of entries to search for (users, groups, or both)
then click Search.
4 02
Chapter 11. Organizing Entries with Groups, Roles, and Views
6. Select the members from the returned entries, and click OK.
7. Click Languages in the left pane to add language-specific information for the group.
4 03
Red Hat D irectory Server 8.2 Administration Guide
8. Click OK to create the new group. It appears in the right pane.
T o edit a static group, double-click the group entry, and make the changes in the editor window. T o view
the changes, go to the View menu, and select Refresh.
NOTE
T he Console for managing static groups may not display all possible selections during a search
operation if there is no VLV index for users' search. T his problem occurs only when the number
of users is 1000 or more and there is no VLV index for search. T o work around the problem,
create a VLV index for the users suffix with the filter (objectclass=person) and scope subtree.
11.1.2. Creating Dynamic Groups in the Console
Dynamic groups filter users based on their DN and include them in a single group.
1. In the Directory Server Console, select the Directory tab.
2. In the left pane, right-click the entry under which to add a new group, and select New > Group.
4 04
Chapter 11. Organizing Entries with Groups, Roles, and Views
Alternatively, go to the Object menu, and select New > Group.
3. Click General in the left pane. T ype a name for the new group in the Group Nam e field (the
name is required), and enter a description of the new group in the Description field.
4. Click Mem bers in the left pane. In the right pane, select the Dynam ic Group tab. Click Add to
create a LDAP URL for querying the database.
5. Enter an LDAP URL in the text field or select Construct to be guided through the construction of
an LDAP URL.
4 05
Red Hat D irectory Server 8.2 Administration Guide
T he results show the current entries (group members) which correspond to the filter.
6. Click Languages in the left pane to add language-specific information for the group.
4 06
Chapter 11. Organizing Entries with Groups, Roles, and Views
7. Click OK. T he new group appears in the right pane.
T o edit a dynamic group, double-click the group entry to open the editor window, and make whatever
changes to the dynamic group. T o view the changes to the group, go to the View menu, and select
Refresh.
NOTE
T he Console for managing dynamic groups may not display all possible selections during a
search operation if there is no VLV index for users' search. T his problem can occur when the
number of users is 1000 or more and there is no VLV index for search. T o work around the
problem, create a VLV index for the users suffix with the filter (objectclass=person) and
scope sub-tree.
11.1.3. Creating Groups in the Command Line
Creating both static and dynamic groups from the command line is a similar process. A group entry
contains the group name, the type of group, and a members attribute.
T here are several different options for the type of group; these are described in more detail in the
Schema Reference. T he type of group in this case refers to the type of defining member attribute it has:
groupOfNam es is a simple group, that allows any entry to be added. T he attribute used to
determine members for this is member.
groupOfUniqueNam es, like groupOfNam es, simply lists user DNs as members, but the members
must be unique. T his prevents users being added more than once as a group member, which is one
4 07
Red Hat D irectory Server 8.2 Administration Guide
way of preventing self-referential group memberships. T he attribute used to determine members for
this is uniqueMember.
groupOfURLs uses a list of LDAP URLs to filter and generate its membership list. T his object class
is required for any dynamic group and can be used in conjunction with groupOfNam es and
groupOfUniqueNam es.
groupOfCertificates is similar to groupOfURLs in that it uses an LDAP filter to search for and
identify certificates (or, really, certificate names) to identify group members. T his is useful for groupbased access control, since the group can be given special access permissions. T he attribute used
to determine members for this is memberCertificate.
T able 11.1, “Dynamic and Static Group Schema” lists the default attributes for groups as they are
created from the command line.
T able 11.1. Dynamic and Static Group Schema
T ype of Group
Group Object Classes
Member Attributes
Static
groupOfUniqueNames
uniqueMember
Dynamic
groupOfUniqueNames
memberURL
groupOfURLs
A static group entry lists the specific members of the group. For example:
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=static group,ou=Groups,dc=example,dc=com
objectClass: top
objectClass: groupofuniquenames
cn: static group
description: Example static group.
uniqueMember: uid=mwhite,ou=People,dc=example,dc=com
uniqueMember: uid=awhite,ou=People,dc=example,dc=com
A dynamic group uses at least one LDAP URL to identify entries belonging to the group and can specify
multiple LDAP URLs or, if used with another group object class like groupOfUniqueNam es, can
explicitly list some group members along with the dynamic LDAP URL.
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=dynamic group,ou=Groups,dc=example,dc=com
objectClass: top
objectClass: groupofuniquenames
objectClass: groupofurls
cn: dynamic group
description: Example dynamic group.
memberURL: ldap:///dc=example,dc=com??sub?(&(objectclass=person)(cn=*sen*))
11.1.4. Using the memberOf Attribute to Manage Group Membership Information
T he entries which belong to a group are defined, in some way, in the group entry itself. T his makes it
very easy to look at a group and see its members and to manage group membership centrally. However,
4 08
Chapter 11. Organizing Entries with Groups, Roles, and Views
there is no good way to find out what groups a single user belongs to. T here is nothing in a user entry
which indicates its memberships, as there are with roles.
T he MemberOf Plug-in correlates group membership lists to the corresponding user entries. An instance
of the MemberOf Plug-in analyzes the member attribute (by default, member) in a group entry and
automatically writes a corresponding memberOf attribute in the member's entry. As membership
changes, the plug-in updates the memberOf attributes on the user entries. T he MemberOf Plug-in
provides a way to view the groups to which a user belongs simply by looking at the entry, including
nested group membership. It can be very difficult to backtrack memberships through nested groups, but
the MemberOf Plug-in shows memberships for all groups, direct and indirect.
NOTE
T he most common people object classes — such as inetorgperson and person — do not
allow the memberof attribute. T o allow the MemberOf Plug-in to add the memberof attribute to a
user entry, make sure that that entry belongs to the inetUser object class, which does allow the
memberof attribute.
Similarly, if an attribute other than memberof is used in the user entry, then make sure that the
user entry belongs to an object class that allows that attribute.
IMPORTANT
T he memberOf attributes for user entries should not be replicated in multi-master environments.
Make sure that the memberOf attribute is excluded from replication in the replication agreement.
(Fractional replication is described in Section 9.1.7, “Replicating a Subset of Attributes with
Fractional Replication”.)
Each server must maintain its own MemberOf Plug-in independently. T o make sure that the
memberOf attributes for entries are the same across servers, simply configure the MemberOf
Plug-in the same on all servers.
With single-master replication, it is perfectly safe to replicate memberOf attributes. Configure the
MemberOf Plug-in for the supplier, then replicate the memberOf attributes to the consumers.
T he MemberOf Plug-in manages member attributes for static groups, not dynamic groups or circular
groups.
11.1.4 .1. T he MemberOf Plug-in Syntax
T he MemberOf Plug-in instance defines two attributes, one for the group member attribute to poll
(memberofgroupattr) and the other for the attribute to create and manage in the member's user entry
(memberofattr).
T he plug-in instance also gives the plug-in path and function to identify the MemberOf Plug-in and
contains a state setting to enable the plug-in, both of which are required for all plug-ins. T he default
MemberOf Plug-in is shown in Example 11.1, “Default MemberOf Plug-in Entry” and the different
parameters are described in T able 11.2, “MemberOf Syntax”.
4 09
Red Hat D irectory Server 8.2 Administration Guide
Example 11.1. Default MemberOf Plug-in Entry
dn: cn=MemberOf Plugin,cn=plugins,cn=config
objectClass: top
objectClass: nsSlapdPlugin
objectClass: extensibleObject
cn: MemberOf Plugin
nsslapd-pluginPath: libmemberof-plugin
nsslapd-pluginInitfunc: memberof_postop_init
nsslapd-pluginType: postoperation
nsslapd-pluginEnabled: on
nsslapd-plugin-depends-on-type: database
memberofgroupattr: member
memberofattr: memberOf
nsslapd-pluginId: memberof
nsslapd-pluginVersion: 8.2.4
nsslapd-pluginVendor: Red Hat, Inc.
nsslapd-pluginDescription: memberof plugin
T able 11.2. MemberOf Syntax
Plug-in Attribute
Description
cn
Gives a unique name for the plug-in instance.
memberofgroupattr
Gives the attribute in the group entry to poll to
identify member DNs. By default, this is the
member attribute, but it can be any attribute used
to identify group members, such as
uniqueMember.
memberofattr
Gives the attribute for the plug-in to create and
manage on the user entry. By default, this is the
memberOf attribute.
nsslapd-pluginEnabled
Sets whether the plug-in instance is enabled
(active) or disabled. T he default MemberOf Plugin instance is disabled by default.
nsslapd-pluginPath
Gives the name of the specific plug-in to load.
T his must be libm em berof-plugin.
nsslapd-pluginInitfunc
Gives the name of the function to call to initialize
the plug-in. T his must be
m em berof_postop_init.
11.1.4 .2. Configuring an Instance of the MemberOf Plug-in from the Command Line
T he attributes defined in the MemberOf Plug-in can be changed, depending on the types of groups used
in the directory.
11.1.4 .2.1. Editing the MemberOf Plug-in from the Console
1. Select the Configuration tab, and expand to the Plugins folder.
2. Scroll to the Mem berof Plugin entry.
4 10
Chapter 11. Organizing Entries with Groups, Roles, and Views
3. Make sure that the plug-in is enabled. T his is disabled by default.
4. Click the Advanced button to open the Advanced Properties Editor.
5. T he memberofgroupattr attribute sets the attribute in the group entry which the server uses to
identify member entries. T he memberofattr attribute sets the attribute which the plug-in creates
and manages on user entries.
6. Save the changes.
7. Restart the server to update the plug-in. For example, open the T asks tab and click the Restart
server task.
4 11
Red Hat D irectory Server 8.2 Administration Guide
11.1.4 .2.2. Editing the MemberOf Plug-in from the Command Line
1. Enable the MemberOf Plug-in.
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=MemberOf Plugin,cn=plugins,cn=config
changetype: replace
replace: nsslapd-pluginEnabled
nsslapd-pluginEnabled: on
-
2. Set the attribute to use for the group member entry attribute. For example:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=MemberOf Plugin,cn=plugins,cn=config
changetype: replace
replace: memberofgroupattr
memberofgroupattr: uniqueMember
-
3. Set the attribute to set on the user entries to show group membership. For example:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=MemberOf Plugin,cn=plugins,cn=config
changetype: replace
replace: memberofattr
memberofattr: memberOf
-
4. Restart the Directory Server to load the modified new plug-in instance.
service dirsrv restart instance_name
11.1.4 .3. Synchronizing memberOf Values
T he MemberOf Plug-in automatically manages the memberOf attribute on group member entries, based
on the configuration in the group entry itself. However, the memberOf attribute can be edited on a user
entry directly (which is improper) or new entries can be imported or replicated over to the server that
have a memberOf attribute already set. T hese situations create inconsistencies between the memberOf
configuration managed by the server plug-in and the actual memberships defined for an entry.
Directory Server has a memberOf repair task which manually runs the plug-in to make sure the
appropriate memberOf attributes are set on entries. T here are three ways to trigger this task:
In the Directory Server Console
Using the fixup-m em berof.pl script
Running a cn=m em berof task,cn=tasks,cn=config tasks entry
4 12
Chapter 11. Organizing Entries with Groups, Roles, and Views
NOTE
T he memberOf regeneration tasks are run locally, even if the entries themselves are replicated.
T his means that the memberOf attributes for the entries on other servers are not updated until
the updated entry is replicated.
11.1.4 .3.1. Initializing and Regenerating memberOf Attributes Using fixup-memberof.pl
T he fixup-m em berof.pl script launches a special task to regenerate all of the memberOf attributes
on user entries based on the member attributes in the group entries. T his is a clean-up task which
synchronizes the membership defined in group entries and the corresponding user entries and
overwrites any accidental or improper edits on the user entries.
1. Open the tool directory for the Directory Server instance,
/usr/lib/dirsrv/slapd-instance_name/.
2. Run the script, binding as the Directory Manager.
./fixup-memberof.pl -D "cn=Directory Manager" -w password
T he fixup-m em berof.pl command is described in more detail in the Configuration and CommandLine Tool Reference.
11.1.4 .3.2. Initializing and Regenerating memberOf Attributes Using ldapmodify
Regenerating memberOf attributes is one of the tasks which can be managed through a special task
configuration entry. T ask entries occur under the cn=tasks configuration entry in the dse.ldif file, so
it is also possible to initiate a task by adding the entry using ldapm odify. As soon as the task is
complete, the entry is removed from the directory.
T he fixup-m em berof.pl script creates a special task entry in a Directory Server instance which
regenerates the memberOf attributes.
T o initiate a memberOf fixup task, add an entry under the cn=m em berof task,
cn=tasks,cn=config entry. T he only required attribute is the cn for the specific task.
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=example memberof,cn=memberof task,cn=tasks,cn=config
cn:example memberof
As soon as the task is completed, the entry is deleted from the dse.ldif configuration, so it is possible
to reuse the same task entry continually.
T he cn=m em berof task configuration is described in more detail in the Configuration and CommandLine Tool Reference.
11.2. Using Roles
Roles are an entry grouping mechanism that unify the static and dynamic groups described in the
previous sections. Roles are designed to be more efficient and easier to use for applications. For
example, an application can get the list of roles of which an entry is a member by querying the entry
itself, rather than selecting a group and browsing the members list of several groups.
4 13
Red Hat D irectory Server 8.2 Administration Guide
itself, rather than selecting a group and browsing the members list of several groups.
11.2.1. About Roles
Red Hat has two kinds of groups. Static groups have a finite and defined list of members. Dynamic
groups used filters to recognize which entries are members of the group, so the group membership is
constantly changed as the entries which match the group filter change. (Both kinds of groups are
described in Section 11.1, “Using Groups”.)
Roles are a sort of hybrid group, behaving as both a static and a dynamic group. With a group, entries
are added to a group entry as members. With a role, the role attribute is added to an entry and then that
attribute is used to identify members in the role entry automatically.
Role members are entries that possess the role. Members can be specified either explicitly or
dynamically. How role membership is specified depends upon the type of role. Directory Server supports
three types of roles:
Managed roles have an explicit enumerated list of members.
Filtered roles are assigned entries to the role depending upon the attribute contained by each entry,
specified in an LDAP filter. Entries that match the filter possess the role.
Nested roles are roles that contain other roles.
Managed roles can do everything that can normally be done with static groups. T he role members can
be filtered using filtered roles, similarly to the filtering with dynamic groups. Roles are easier to use than
groups, more flexible in their implementation, and reduce client complexity.
When a role is created, determine whether a user can add themselves or remove themselves from the
role. See Section 11.2.10, “Using Roles Securely” for more information about roles and access control.
NOTE
Evaluating roles is more resource-intensive for the Directory Server than evaluating groups
because the server does the work for the client application. With roles, the client application can
check role membership by searching for the nsRole attribute. T he nsRole attribute is a computed
attribute, which identifies to which roles an entry belongs; the nsRole attribute is not stored with
the entry itself. From the client application point of view, the method for checking membership is
uniform and is performed on the server side.
Considerations for using roles are covered in the Directory Server Deployment Guide.
11.2.2. Creating a Managed Role
Managed roles have an explicit enumerated list of members. Managed roles are added to entries by
adding the nsRoleDN attribute to the entry.
11.2.2.1. Creating a Managed Role in the Console
1. In the Directory Server Console, select the Directory tab.
2. Browse the tree in the left navigation pane, and select the parent entry for the new role.
3. Go to the Object menu, and select New > Role.
4 14
Chapter 11. Organizing Entries with Groups, Roles, and Views
Alternatively, right-click the entry and select New > Role.
4. Click General in the left pane. T ype a name for the new role in the Role Nam e field. T he role
name is required.
5. Enter a description of the new role in the Description field.
6. Click Mem bers in the left pane.
7. In the right pane, select Managed Role. Click Add to add new entries to the list of members.
8. In the Search drop-down list, select Users from the Search drop-down list, then click Search.
4 15
Red Hat D irectory Server 8.2 Administration Guide
Select one of the entries returned, and click OK.
9. After adding all of the entries, click OK.
11.2.2.2. Creating Managed Roles through the Command Line
Roles inherit from the ldapsubentry object class, which is defined in the IT U X.509 standard. In
addition, each managed role requires two object classes that inherit from the nsRoleDefinition
object class:
nsSimpleRoleDefinition
nsManagedRoleDefinition
A managed role also allows an optional description attribute.
Members of a managed role have the nsRoleDN attribute in their entry.
T his example creates a role which can be assigned to the marketing department.
1. Use ldapm odify with the -a option to add the managed role entry. T he new entry must contain
the nsManagedRoleDefinition object class, which in turn inherits from the LdapSubEntry,
nsRoleDefinition, and nsSim pleRoleDefinition object classes.
ldapmodify -a -D "cn=Directory Manager" -w secret -h host -p 389
dn: cn=Marketing,ou=people,dc=example,dc=com
objectclass: top
objectclass: LdapSubEntry
objectclass: nsRoleDefinition
objectclass: nsSimpleRoleDefinition
objectclass: nsManagedRoleDefinition
cn: Marketing
description: managed role for marketing staff
2. Assign the role to the marketing staff members, one by one, using ldapm odify:
4 16
Chapter 11. Organizing Entries with Groups, Roles, and Views
ldapmodify -D "cn=Directory Manager" -w secret -h host -p 389
dn: cn=Bob,ou=people,dc=example,dc=com
changetype: modify
add: nsRoleDN
nsRoleDN: cn=Marketing,ou=people,dc=example,dc=com
T he nsRoleDN attribute in the entry indicates that the entry is a member of a managed role,
cn=Marketing,ou=people,dc=exam ple,dc=com .
11.2.3. Creating a Filtered Role
Entries are assigned to a filtered role depending whether the entry possesses a specific attribute
defined in the role. T he role definition specifies an LDAP filter for the target attributes. Entries that match
the filter possess (are members of) the role.
11.2.3.1. Creating a Filtered Role in the Console
1. In the Directory Server Console, select the Directory tab.
2. Browse the tree in the left navigation pane, and select the parent entry for the new role.
3. Go to the Object menu, and select New > Role.
Alternatively, right-click the entry and select New > Role.
4. Click General in the left pane. T ype a name for the new role in the Role Nam e field. T he role
name is required.
4 17
Red Hat D irectory Server 8.2 Administration Guide
5. Enter a description of the new role in the Description field.
6. Click Mem bers in the left pane.
A search dialog box appears briefly.
7. In the right pane, select Filtered Role.
8. Enter an LDAP filter in the text field, or click Construct to be guided through the construction of
an LDAP filter.
T he Construct opens the standard LDAP URL construction dialog. Ignore the fields for LDAP
Server Host, Port, Base DN, and Search (since the search scope cannot be set filtered role
definitions).
Select the types of entries to filter from the For drop-down list. T he entries can be users,
groups, or both.
Select an attribute from the Where drop-down list. T he two fields following it refine the search
by selecting one of the qualifiers from the drop-down list, such as contains, does not
contain, is, or is not. Enter an attribute value in the text box. T o add additional filters, click
More. T o remove unnecessary filters, click Fewer.
9. Click T est to try the filter.
4 18
Chapter 11. Organizing Entries with Groups, Roles, and Views
10. Click OK.
11.2.3.2. Creating a Filtered Role through the Command Line
Roles inherit from the ldapsubentry object class, which is defined in the IT U X.509 standard. In
addition, each filtered role requires two object classes that inherit from the nsRoleDefinition object
class:
nsComplexRoleDefinition
nsFilteredRoleDefinition
A filtered role entry also requires the nsRoleFilter attribute to define the LDAP filter to determine role
members. Optionally, the role can take a description attribute.
Members of a filtered role are entries that match the filter specified in the nsRoleFilter attribute.
T his example creates a filtered role which is applied to all sales managers.
1. Run ldapm odify with the -a option to add a new entry:
ldapmodify -a -D "cn=Directory Manager" -w secret -h host -p 389
4 19
Red Hat D irectory Server 8.2 Administration Guide
2. Create the filtered role entry.
T he role entry has the nsFilteredRoleDefinition object class, which inherits from the
LdapSubEntry, nsRoleDefinition, and nsCom plexRoleDefinition object classes.
T he nsRoleFilter attribute sets a filter for o (organization) attributes that contain a value of
sales m anagers.
dn: cn=SalesManagerFilter,ou=people,dc=example,dc=com
objectclass: top
objectclass: LDAPsubentry
objectclass: nsRoleDefinition
objectclass: nsComplexRoleDefinition
objectclass: nsFilteredRoleDefinition
cn: SalesManagerFilter
nsRoleFilter: o=sales managers
Description: filtered role for sales managers
T he following entry matches the filter (possesses the o attribute with the value sales m anagers), and,
therefore, it is a member of this filtered role automatically:
dn: cn=Pat Smith,ou=people,dc=example,dc=com
objectclass: person
cn: Pat
sn: Smith
userPassword: secret
o: sales managers
11.2.4. Creating a Nested Role
Nested roles are roles that contain other roles. Before it is possible to create a nested role, another role
must exist. When a nested role is created, the Console displays a list of the roles available for nesting.
T he roles nested within the nested role are specified using the nsRoleDN attribute.
11.2.4 .1. Creating a Nested Role in the Console
1. In the Directory Server Console, select the Directory tab.
2. Browse the tree in the left navigation pane, and select the parent entry for the new role.
3. Go to the Object menu, and select New > Role.
4 20
Chapter 11. Organizing Entries with Groups, Roles, and Views
Alternatively, right-click the entry and select New > Role.
4. Click General in the left pane. T ype a name for the new role in the Role Nam e field. T he role
name is required.
5. Click Mem bers in the left pane.
6. In the right pane, select Nested Role.
7. Click Add to add roles to the list. T he members of the nested role are members of other existing
roles.
4 21
Red Hat D irectory Server 8.2 Administration Guide
8. Select a role from the Available roles list, and click OK.
11.2.4 .2. Creating Nested Role through the Command Line
Roles inherit from the ldapsubentry object class, which is defined in the IT U X.509 standard. In
addition, each nested role requires two object classes that inherit from the nsRoleDefinition object
class:
nsComplexRoleDefinition
nsNestedRoleDefinition
A nested role entry also requires the nsRoleDN attribute to identify the roles to nest within the container
role. Optionally, the role can take a description attribute.
Members of a nested role are members of the roles specified in the nsRoleDN attributes of the nested
role definition entry.
T his example creates a single role out of the managed marketing role and filtered sales manager role.
1. Run ldapm odify with the -a option to add a new entry:
ldapmodify -a -D "cn=Directory Manager" -w secret -h host -p 389
2. Create the nested role entry. T he nested role has four object classes:
nsNestedRoleDefinition
LDAPsubentry (inherited)
nsRoleDefinition (inherited)
nsCom plexRoleDefinition (inherited)
T he nsRoleDN attributes contain the DNs for both the marketing managed role and the sales
managers filtered role.
dn: cn=MarketingSales,ou=people,dc=example,dc=com
objectclass: top
objectclass: LDAPsubentry
objectclass: nsRoleDefinition
objectclass: nsComplexRoleDefinition
objectclass: nsNestedRoleDefinition
cn: MarketingSales
nsRoleDN: cn=SalesManagerFilter,ou=people,dc=example,dc=com
nsRoleDN: cn=Marketing,ou=people,dc=example,dc=com
Both of the users in the previous examples, Bob and Pat, are members of this new nested role.
4 22
Chapter 11. Organizing Entries with Groups, Roles, and Views
11.2.5. Editing and Assigning Roles to an Entry
T he entries which belong to a role are assigned on the role entry itself. For managed roles, user entries
are added explicitly; for filtered roles, they are added through the results of an LDAP filter.
User entries are assigned to the role through the command line by editing the role entry, either by
adding the entry as a member or adjusting the filter so it is included. In the Directory Server Console,
however, there is a shortcut to add entries to a role by apparently editing the desired user entry (but,
functionally, this really edits the role entry).
1. Select the Directory tab.
2. In the left navigation pane, browse the tree, and select the entry for which to view or edit a role.
3. Select Set Roles from the Object menu.
4. Select the Managed Roles tab to display the managed roles to which this entry belongs.
5. T o add a new managed role, click Add, and select an available role from the Role Selector
window.
4 23
Red Hat D irectory Server 8.2 Administration Guide
TIP
T he configuration for a managed role associated with an entry can be edited by clicking the
Edit button. T he Edit Entry dialog box opens, and the general information or members
for the role can be changed.
6. Select the Other Roles tab to view the filtered or nested roles to which this entry belongs.
Click Edit to make changes to any filtered or nested roles associated with the entry.
11.2.6. Viewing Roles for an Entry through the Command Line
Role assignments are always visible for an entry when it is displayed in the Directory Server Console.
Role assignments are not returned automatically through the command line, however.
T he nsRole attribute is an operational attribute. In LDAP, operational attributes must be requested
explicitly in the search attributes list; they are not returned by default with the regular attributes in the
schema of the entry. For example, this ldapsearch command returns the list of roles of which
uid=scarter is a member, in addition to the regular attributes for the entry:
/usr/lib64/mozldap/ldapsearch -D "cn=directory manager" -w secret -p 389 -h
server.example.com “(uid=scarter)” \* nsRole
dn: uid=scarter,ou=people,dc=example,dc=com
objectClass: inetorgperson
objectClass: top
objectClass: person
objectClass: organizationalperson
uid: scarter
cn: Sam Carter
sn: Carter
givenName: Sam
mail: [email protected]
userPassword: {SSHA}6BE31mhTfcYyIQF60kWlnEL8sIvPZ59hvFTRKw==
manager: uid=lbrown,ou=people,dc=example,dc=com
nsrole: cn=Role for Managers,dc=example,dc=com
nsrole: cn=Role for Accounting,dc=example,dc=com
4 24
Chapter 11. Organizing Entries with Groups, Roles, and Views
IMPORTANT
Be sure to use the nsRole attribute, not the nsRoleDN attribute, to evaluate role membership.
11.2.7. Making a Role Inactive or Active
T he concept of activating/inactivating roles allows entire groups of entries to be activated or inactivated
in just one operation. T hat is, the members of a role can be temporarily disabled by inactivating the role
to which they belong.
When a role is inactivated, it does not mean that the user cannot bind to the server using that role entry.
T he meaning of an inactivated role is that the user cannot bind to the server using any of the entries
that belong to that role; the entries that belong to an inactivated role will have the nsAccountLock
attribute set to true.
Members of a role can be temporarily disabled by inactivating the role to which they belong. Inactivating a
role inactivates the entries possessed by the role, not the role itself.
1. Select the Directory tab.
2. Browse the navigation tree in the left pane to locate the base DN for the role. Roles appear in the
right pane with other entries.
3. Double-click the role, open the Account tab, and click the Inactivate button.
Alternatively, select the role. Right-click the role and select Inactivate from the menu.
T he role is inactivated.
T o reactivate a disabled role, re-open the role configuration or open the Object menu, and select
Activate. All of the members of the role are re-enabled.
11.2.8. Viewing the Activation Status for Entries
When a nested role is inactivated, a user cannot bind to the server if it is a member of any role within the
nested role. All the entries that belong to a role that directly or indirectly are members of the nested role
have nsAccountLock set to true. T here can be several layers of nested roles, and inactivating a
4 25
Red Hat D irectory Server 8.2 Administration Guide
nested role at any point in the nesting will inactivate all roles and users beneath it.
T he Directory Server Console automatically shows the active or inactive status of entries.
T o see the inactivated entries, select Inactivation State from the View menu. Members of an
inactivated role have a red slash through them. For example, John Smith is a member of the inactive
Example Role.
T he nsAccountLock attribute is an operational attribute and must be explicitly requested in the search
command in the list of search attributes. For example:
/usr/lib64/mozldap/ldapsearch -D "cn=directory manager" -w secret -p 389 -h
server.example.com “(uid=scarter)” nsAccountLock
11.2.9. About Deleting Roles
Deleting a role deletes the role entry but does not delete the nsRoleDN attribute for each role member.
T o delete the nsRoleDN attribute for each role member, enable the Referential Integrity plug-in, and
configure it to manage the nsRoleDN attribute. For more information on the Referential Integrity plug-in,
see Section 3.5, “Maintaining Referential Integrity”.
11.2.10. Using Roles Securely
Not every role is suitable for use in a security context. When creating a new role, consider how easily the
role can be assigned to and removed from an entry. Sometimes it is appropriate for users to be able to
add or remove themselves easily from a role. For example, if there is an interest group role called
Mountain Biking, interested users should be able to add themselves or remove themselves easily.
However, it is inappropriate to have such open roles for some security situations. One potential security
risk is inactivating user accounts by inactivating roles. Inactive roles have special ACIs defined for their
suffix. If an administrator allows users to add and remove themselves from roles freely, then in some
circumstance, they may be able to remove themselves from an inactive role to prevent their accounts
from being locked.
For example, user A possesses the managed role, MR. T he MR role has been locked using account
inactivation. T his means that user A cannot bind to the server because the nsAccountLock attribute is
computed as true for that user. However, if user A was already bound to Directory Server and noticed
that he is now locked through the MR role, the user can remove the nsRoleDN attribute from his entry
and unlock himself if there are no ACIs preventing him.
T o prevent users from removing the nsRoleDN attribute, use the following ACIs depending upon the type
of role being used.
Managed roles. For entries that are members of a managed role, use the following ACI to prevent
users from unlocking themselves by removing the appropriate nsRoleDN:
4 26
Chapter 11. Organizing Entries with Groups, Roles, and Views
aci: (targetattr="nsRoleDN") (targattrfilters=
add=nsRoleDN:(!(nsRoleDN=cn=AdministratorRole,dc=example,dc=com)),
del=nsRoleDN:(!(nsRoleDN=cn=nsManagedDisabledRole,dc=example,dc=com)))
(version3.0;aci allow mod of nsRoleDN by self but not to critical values;
allow(write) userdn=ldap:///self;)
Filtered roles. T he attributes that are part of the filter should be protected so that the user cannot
relinquish the filtered role by modifying an attribute. T he user should not be allowed to add, delete, or
modify the attribute used by the filtered role. If the value of the filter attribute is computed, then all
attributes that can modify the value of the filter attribute should be protected in the same way.
Nested roles. A nested role is comprised of filtered and managed roles, so both ACIs should be
considered for modifying the attributes (nsRoleDN or something else) of the roles that comprise the
nested role.
For more information about account inactivation, see Section 13.8, “Inactivating Users and Roles”.
11.3. Using Views
Virtual directory tree views, or views, create a virtual directory hierarchy, so it is easy to navigate entries,
without having to make sure those entries physically exist in any particular place. T he view uses
information about the entries to place them in the view hierarchy, similarly to members of a filtered role or
a dynamic group. Views superimpose a DIT hierarchy over a set of entries, and to client applications,
views appear as ordinary container hierarchies.
Views create a directory tree similar to the regular hierarchy, such as using organizational unit entries
for subtrees, but views entries have an additional object class (nsview) and a filter attribute
(nsviewfilter) that set up a filter for the entries which belong in that view. Once the view container
entry is added, all of the entries that match the view filter instantly populate the view. T he target entries
only appear to exist in the view; their true location never changes. For example, a view may be created
as ou=Location Views, and a filter is set for l=Mountain View. Every entry, such as cn=Jane
Sm ith,l=Mountain View,ou=People,dc=exam ple,dc=com , is immediately listed under the
ou=Location Views entry, but the real cn=Jane Sm ith entry remains in the
ou=People,dc=exam ple,dc=com subtree.
4 27
Red Hat D irectory Server 8.2 Administration Guide
Figure 11.1. A Directory T ree with a Virtual DIT View hierarchy
Virtual DIT views behave like normal DIT s in that a subtree or a one-level search can be performed with
the expected results being returned.
NOTE
T here is a sample LDIF file with example views entries, Exam ple-views.ldif, installed with
Directory Server. T his file is in the /usr/share/dirsrv/data directory on Red Hat Enterprise
Linux 5 (64-bit).
T he Directory Server Deployment Guide has more information on how to integrate views with the
directory tree hierarchy.
11.3.1. Creating Views in the Console
1. Select the Directory tab.
2. In the left navigation tree, create an organizational unit suffix to hold the views. For instance, for
views based on the locality (l) attribute, name this organizational unit Location Views.
Creating sub suffixes is described in Section 2.1.1.2, “Creating a New Sub Suffix Using the
Console”.
3. Right-click ou=Location Views, and select New > Other.
4. Select nsview from the New Object menu, and click OK.
4 28
Chapter 11. Organizing Entries with Groups, Roles, and Views
5. In the Property Editor window, click the Add Value button, and add the organization unit
object class.
6. Name the organization unit according to how to organize the views. For instance, ou=Sunnyvale.
Make the ou attribute the naming attribute.
4 29
Red Hat D irectory Server 8.2 Administration Guide
7. Click the Add Attribute button, and add the nsviewfilter attribute.
8. Create a filter that reflects the views, such as (l=Sunnyvale).
4 30
Chapter 11. Organizing Entries with Groups, Roles, and Views
9. Click the Change button in the lower right corner to change the naming attribute.
Use the ou of the entry as the naming attribute instead of description.
10. Click OK to close the attributes box, and click OK again to save the new view entry.
T he new view is immediately populated with any entries matching the search filter, and any new entries
added to directory are automatically included in the view.
11.3.2. Creating Views from the Command Line
1. Use the ldapm odify utility to bind to the server and prepare it to add the new view entry to the
configuration file.
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389
-h server.example.com
2. Add the new views container entry, in this example, under the dc=exam ple,dc=com root suffix.
4 31
Red Hat D irectory Server 8.2 Administration Guide
T his entry must have the nsview object class and the nsViewFilter attribute. T he
nsViewFilter attribute sets the attribute-value which identifies entries that belong in the view.
dn: ou=Example View,dc=example,dc=com
objectClass: top
objectClass: organizationalunit
objectClass: nsview
ou=Example View
nsViewFilter: l=Mountain View
description: Example View
4 32
Chapter 12. Managing Access Control
Chapter 12. Managing Access Control
Red Hat Directory Server allows you to control access to your directory. T his chapter describes the how
to implement access control. T o take full advantage of the power and flexibility of access control, while
you are in the planning phase for your directory deployment, define an access control strategy as an
integral part of your overall security policy.
12.1. Access Control Principles
T he mechanism which defines user access is called access control. When the server receives a
request, it uses the authentication information provided by the user in the bind operation and the access
control instructions (ACIs) defined in the server to allow or deny access to directory information. T he
server can allow or deny permissions for actions on entries like read, write, search, and compare. T he
permission level granted to a user may depend on the authentication information provided.
Access control in Directory Server is flexible enough to provide very precise rules on when the ACIs are
applicable:
For the entire directory, a subtree of the directory, specific entries in the directory (including entries
defining configuration tasks), or a specific set of entry attributes.
For a specific user, all users belonging to a specific group or role, or all users of the directory.
For a specific location such as an IP address or a DNS name.
12.1.1. ACI Structure
Access control instructions are stored in the directory as attributes of entries. T he aci attribute is an
operational attribute; it is available for use on every entry in the directory, regardless of whether it is
defined for the object class of the entry. It is used by the Directory Server to evaluate what rights are
granted or denied when it receives an LDAP request from a client. T he aci attribute is returned in an
ldapsearch operation if specifically requested.
T he three main parts of an ACI statement are:
T arget
Permission
Bind Rule
T he permission and bind rule portions of the ACI are set as a pair, also called an access control rule
(ACR). T he specified permission is granted or denied depending on whether the accompanying rule is
evaluated to be true.
12.1.2. ACI Placement
If an entry containing an ACI does not have any child entries, the ACI applies to that entry only. If the
entry has child entries, the ACI applies to the entry itself and all entries below it. As a direct
consequence, when the server evaluates access permissions to any given entry, it verifies the ACIs for
every entry between the one requested and the directory suffix, as well as the ACIs on the entry itself.
T he aci attribute is multi-valued, which means that you can define several ACIs for the same entry or
subtree.
An ACI created on an entry can be set so it does not apply directly to that entry but to some or all of the
entries in the subtree below it. T he advantage of this is that general ACIs can be placed at a high level in
the directory tree that effectively apply to entries more likely to be located lower in the tree. For example,
an ACI that targets entries that include the inetorgperson object class can be created at the level of
4 33
Red Hat D irectory Server 8.2 Administration Guide
an organizationalUnit entry or a locality entry.
Minimize the number of ACIs in the directory tree by placing general rules at high level branch points. T o
limit the scope of more specific rules, place them as close as possible to leaf entries.
NOTE
ACIs placed in the root DSE entry apply only to that entry.
12.1.3. ACI Evaluation
T o evaluate the access rights to a particular entry, the server compiles a list of the ACIs present on the
entry itself and on the parent entries back up to the top level entry stored on the Directory Server. ACIs
are evaluated across all of the databases for a particular Directory Server but not across all Directory
Server instances.
T he evaluation of this list of ACIs is done based on the semantics of the ACIs, not on their placement in
the directory tree. T his means that ACIs that are close to the root of the directory tree do not take
precedence over ACIs that are closer to the leaves of the directory tree.
For Directory Server ACIs, the precedence rule is that ACIs that deny access take precedence over ACIs
that allow access. Between ACIs that allow access, union semantics apply, so there is no precedence.
For example, if you deny write permission at the directory's root level, then none of the users can write to
the directory, regardless of the specific permissions you grant them. T o grant a specific user write
permissions to the directory, you have to restrict the scope of the original denial for write permission so
that it does not include the user.
12.1.4. ACI Limitations
When creating an access control policy for your directory service, you need to be aware of the following
restrictions:
If your directory tree is distributed over several servers using the chaining feature, some restrictions
apply to the keywords you can use in access control statements:
ACIs that depend on group entries (groupdn keyword) must be located on the same server as
the group entry. If the group is dynamic, then all members of the group must have an entry on the
server, too. If the group is static, the members' entries can be located on remote servers.
ACIs that depend on role definitions (roledn keyword) must be located on the same server as
the role definition entry. Every entry that is intended to have the role must also be located on the
same server.
However, you can match values stored in the target entry with values stored in the entry of the bind
user; for example, using the userattr keyword. Access is evaluated normally even if the bind user
does not have an entry on the server that holds the ACI.
For more information on how to chain access control evaluation, see Section 2.3.6, “Database Links
and Access Control Evaluation”.
Attributes generated by class of service (CoS) cannot be used in all ACI keywords. Specifically, you
should not use attributes generated by CoS with the following keywords:
targetfilter (Section 12.3.2.4, “T argeting Entries or Attributes Using LDAP Filters”)
targattrfilters (Section 12.3.2.2, “T argeting Attributes”)
userattr (Section 12.4.5.1, “Using the userattr Keyword”)
If you create target filters or bind rules that depend on the value of attributes generated by CoS, the
4 34
Chapter 12. Managing Access Control
access control rule will not work. For more information on CoS, see Chapter 11, Organizing Entries
with Groups, Roles, and Views.
Access control rules are always evaluated on the local server. T herefore, it is not necessary to
specify the hostname or port number of the server in LDAP URLs used in ACI keywords. If you do, the
LDAP URL is not taken into account at all. For more information on LDAP URLs, see Appendix B,
LDAP URLs.
12.2. Default ACIs
When the Admin Server is set up, the following default ACIs apply to the directory information stored in
the userRoot database:
Users can modify a list of common attributes in their own entries, including the mail,
telephoneNumber, userPassword, and seeAlso attributes. Operational and most of the security
attributes, such as aci, nsroledn, and passwordExpirationTime, cannot be modified by users.
Users have anonymous access to the directory for search, compare, and read operations.
T he administrator (by default
uid=adm in,ou=Adm inistrators,ou=T opologyManagem ent,o=NetscapeRoot) has all
rights except proxy rights.
All members of the Configuration Adm inistrators group have all rights except proxy rights.
All members of the Directory Adm inistrators group have all rights except proxy rights.
Server Instance Entry (SIE) group.
T he NetscapeRoot subtree has its own set of default ACIs:
All members of the Configuration Adm inistrators group have all rights on the
NetscapeRoot subtree except proxy rights.
Users have anonymous access to the NetscapeRoot subtree for search and read operations.
All authenticated users have search, compare, and read rights to configuration attributes that identify
the Admin Server.
Group expansion.
T he following sections explain how to modify these default settings.
12.3. Creating ACIs Manually
You can create access control instructions manually using LDIF statements and add them to your
directory tree using the ldapm odify utility, similar to the instructions in Section 3.3, “Using LDIF Update
Statements to Create or Modify Entries”. T he following sections explain in detail how to create the LDIF
statements.
4 35
Red Hat D irectory Server 8.2 Administration Guide
NOTE
LDIF ACI statements can be very complex. However, if you are setting access control for a large
number of directory entries, using LDIF is the preferred because it is faster than using the
Console. T o familiarize yourself with LDIF ACI statements, however, you may want to use the
Directory Server Console to set the ACI and then click the Edit Manually button on the
Access Control Editor. T his shows you the correct LDIF syntax. If your operating system
allows it, you can even copy the LDIF from the Access Control Editor and paste it into your
LDIF file.
12.3.1. The ACI Syntax
T he aci attribute uses the following syntax:
aci: (target)(version 3.0;acl "name";permissionbind_rules;)
target specifies the entry, attributes, or set of entries and attributes for which to control access. T he
target can be a distinguished name, one or more attributes, or a single LDAP filter. T he target is an
optional part of the ACI.
version 3.0 is a required string that identifies the ACI version.
name is a name for the ACI. T he name can be any string that identifies the ACI. T he ACI name is
required.
permission specifically outlines what rights are being allowed or denied; for example, read or search
rights.
bind_rules specify the credentials and bind parameters that a user has to provide to be granted
access. Bind rules can also specifically deny access to certain users or groups of users.
You can have multiple permission-bind rule pairs for each target. T his allows you to set multiple access
controls for a given target efficiently. For example:
target(permissionbind_rule)(permissionbind_rule)...
If you have several ACRs in one ACI statement, the syntax is in the following form:
aci: (target)(version 3.0;acl "name";permissionbind_rule; permissionbind_rule; ...
permissionbind_rule;)
T he following is an example of a complete LDIF ACI:
aci: (target="ldap:///uid=bjensen,dc=example,dc=com")(targetattr=*)
(version 3.0;acl "aci1";allow (write) userdn="ldap:///self";)
In this example, the ACI states that the user bjensen has rights to modify all attributes in her own
directory entry.
12.3.2. Defining Targets
T he target identifies to what the ACI applies. If the target is not specified, the ACI applies to the entry
containing the aci attribute and to the entries below it. A target can be any of the following:
A directory entry or all of the entries in a subtree, as described in Section 12.3.2.1, “T argeting a
4 36
Chapter 12. Managing Access Control
Directory Entry”.
Attributes of an entry, as described in Section 12.3.2.2, “T argeting Attributes”.
A set of entries or attributes that match a specified LDAP filter, as described in Section 12.3.2.4,
“T argeting Entries or Attributes Using LDAP Filters”.
An attribute value, or a combination of values, that match a specified LDAP filter, as described in
Section 12.3.2.5, “T argeting Attribute Values Using LDAP Filters”.
T he general syntax for a target is as follows:
(keyword = "expression")
(keyword != "expression")
keyword indicates the type of target.
equal (=) indicates that the target is the object specified in the expression, and not equal (!=)
indicates the target is not the object specified in the expression.
expression identifies the target.
T he quotation marks ("") around expression are required. What you use for expression is dependent
upon the keyword that you supply.
T able 12.1, “LDIF T arget Keywords” lists each keyword and the associated expressions.
T able 12.1. LDIF T arget Keywords
Keyword
Valid Expressions
Wildcard Allowed
target
ldap:///distinguished_name
Yes
targetattr
attribute
Yes
targetfilter
LDAP_filter
Yes
targetattrfilters
LDAP_operation:LDAP_filter
Yes
In all cases, you must keep in mind that when you place an ACI on an entry, if it is not a leaf entry, the
ACI also applies to all entries below it. For example, if you target the entry
ou=accounting,dc=exam ple,dc=com , the permissions you set apply to all entries in the
accounting branch of the exam ple tree.
As a counter example, if you place an ACI on the ou=accounting,dc=exam ple,dc=com entry, you
cannot target the uid=sarette,ou=people,dc=exam ple,dc=com entry because it is not located
under the accounting tree.
Be wary of using != when specifying an attribute to deny. ACLs are treated as a logical OR, which
means that if you created two ACLs as shown below, the result allows all values of the target attribute.
acl1: ( target=...)( targetattr!=a )(version 3.0; acl "name";allow (...)..
acl2: ( target=...)( targetattr!=b )(version 3.0; acl "name";allow (...)..
T he first ACL (acl1) allows b and the second ACL (acl2) allows a. T he result of these two ACLs is the
same as the one resulting from using an ACL of the following form:
acl3: ( targetattr="*" ) allow (...) ...
In the second example, nothing is denied, which could give rise to security problems.
4 37
Red Hat D irectory Server 8.2 Administration Guide
When you want to deny access to a particular attribute, use deny in the permissions clause rather than
using allow with ( targetattr != value ). For example, usages such as these are
recommended:
acl1: ( target=...)( targetattr=a )(version 3.0; acl "name";deny (...)..
acl2: ( target=...)( targetattr=b )(version 3.0; acl "name";deny (...)..
12.3.2.1. T argeting a Directory Entry
T o target a directory entry (and the entries below it), you must use the target keyword. T he target
keyword can accept a value of the following format:
target="ldap:///distinguished_name
T his identifies the distinguished name of the entry to which the access control rule applies. For example:
(target = "ldap:///uid=bjensen,dc=example,dc=com")
NOTE
If the DN of the entry to which the access control rule applies contains a comma, escape the
comma with a single backslash (\), such as
(target="ldap:///uid=lfuentes,dc=exam ple.com Bolivia\,S.A.").
Wildcards can be used when targeting a distinguished name using the target keyword. T he wildcard
indicates that any character or string or substring is a match for the wildcard. Pattern matching is based
on any other strings that have been specified with the wildcard.
T he following are legal examples of wildcard usage:
(target="ldap:///uid=* ,dc=exam ple,dc=com ") — Matches every entry in the entire
exam ple tree that has the uid attribute in the entry's RDN.
(target="ldap:///uid=* Anderson,dc=exam ple,dc=com ") — Matches every entry directly
under the exam ple node with a uid ending in Anderson.
(target="ldap:///uid=C* A,dc=exam ple,dc=com ") — Matches every entry directly under
the exam ple node with a uid beginning with C and ending with A.
(target="ldap:///uid=* ,dc=exam ple,dc=com ") — Matches every entry in the entire
exam ple tree that has the uid attribute in the entry's RDN.
(target="ldap:///uid=* ,ou=* ,dc=exam ple,dc=com ") — Matches every entry in the
exam ple tree whose distinguished name contains the uid and ou attributes. T hus,
uid=fchen,ou=Engineering,dc=exam ple,dc=com or
uid=claire,ou=Engineering,ou=people,dc=exam ple,dc=com would match, but
uid=bjensen,dc=exam ple,dc=com ou=Engineering,dc=exam ple,dc=com would not.
Depending on the position of the wildcard, it can apply to the full DN, not only to attribute values.
T herefore, the wildcard can be used as a substitute for portions of the DN. For example,
uid=andy* ,dc=exam ple,dc=com targets all the directory entries in the entire exam ple tree with a
matching uid attribute and not just the entries that are immediately below the dc=exam ple,dc=com
node. In other words, this target matches with longer expressions such as
uid=andy,ou=eng,dc=exam ple,dc=com or uid=andy,ou=m arketing,dc=exam ple,dc=com .
4 38
Chapter 12. Managing Access Control
NOTE
You cannot use wildcards in the suffix part of a distinguished name. T hat is, if your directory uses
the suffixes c=US and c=GB, then you cannot use (target="ldap:///dc=exam ple,c=* ")
as a target to reference both suffixes. Neither can you use a target such as
uid=bjensen,dc=* .com .
12.3.2.2. T argeting Attributes
In addition to targeting directory entries, you can also target one or more attributes included in the
targeted entries. T his is useful to deny or allow access to partial information about an entry. For
example, you could allow access to only the common name, surname, and telephone number attributes
of a given entry while denying access to sensitive information such as passwords.
You can specify that the target is equal or is not equal to a specific attribute. T he attributes you supply
do not need to be defined in the schema. T his absence of schema checking makes it possible to
implement an access control policy when you set up your directory service for the first time, even if the
ACLs you create do not apply to the current directory content.
T o target attributes, use the targetattr keyword. T he keyword uses the following syntax:
(targetattr = "attribute")
You can target multiple attributes by using the targetattr keyword with the following syntax:
(targetattr = "attribute1 || attribute2 ...|| attributen")
attributeX is the name of the targeted attribute. For example, this targets the common name (cn)
attribute:
(targetattr = "cn")
T o target an entry's common name, surname, and UID attributes, use the following:
(targetattr = "cn || sn || uid")
T he attributes specified in the targetattr keyword apply to the entry that the ACI is targeting and to
all the entries below it. If you target the password attribute on the entry
uid=bjensen,ou=Marketing,dc=exam ple,dc=com , only the password attribute on the bjensen
entry is affected by the ACI because it is a leaf entry.
If, however, you target the tree's branch point ou=Marketing,dc=exam ple,dc=com , then all the
entries beneath the branch point that can contain a password attribute are affected by the ACI.
12.3.2.3. T argeting Both an Entry and Attributes
By default, the entry targeted by an ACI containing a targetattr keyword is the entry on which the ACI
is placed. T hat is, putting an ACI such as aci: (targetattr = "uid")(access_control_rules;) on
the ou=Marketing,dc=exam ple,dc=com entry means that the ACI applies to the entire Marketing
subtree. However, you can also explicitly specify a target using the target keyword:
4 39
Red Hat D irectory Server 8.2 Administration Guide
aci:
(target="ldap:///ou=Marketing,dc=example,dc=com")(targetattr="uid")(access_control
_rules;)
T he order in which you specify the target and the targetattr keywords is not important.
12.3.2.4 . T argeting Entries or Attributes Using LDAP Filters
You can use LDAP filters to target a group of entries that match certain criteria. T o do this, you must use
the targetfilter keyword with an LDAP filter. T he syntax of the targetfilter keyword is as
follows:
(targetfilter = "LDAP_filter")
LDAP_filter is a standard LDAP search filter. For more information on the syntax of LDAP search filters,
see Chapter 8, Finding Directory Entries.
For example, suppose that all entries in the accounting department include the attribute-value pair
ou=accounting, and all entries in the engineering department include the attribute-value pair
ou=engineering subtree. T he following filter targets all the entries in the accounting and engineering
branches of the directory tree:
(targetfilter = "(|(ou=accounting)(ou=engineering))")
T his type of filter targets whole entries. You can associate the targetfilter and the targetattr
keywords to create ACIs that apply to a subset of attributes in the targeted entries.
T he following LDIF example allows members of the Engineering Admins group to modify the
departmentNumber and manager attributes of all entries in the Engineering business category. T his
example uses LDAP filtering to select all entries with businessCategory attributes set to
Engineering:
dn: dc=example,dc=com
objectClass: top
objectClass: organization
aci: (targetattr="departmentNumber || manager")
(targetfilter="(businessCategory=Engineering)")
(version 3.0; acl "eng-admins-write"; allow (write)
groupdn ="ldap:///cn=Engineering Admins,dc=example,dc=com";)
NOTE
Although using LDAP filters can be useful when you are targeting entries and attributes that are
spread across the directory, the results are sometimes unpredictable because filters do not
directly name the object for which you are managing access. T he set of entries targeted by a
filtered ACI is likely to change as attributes are added or deleted. T herefore, if you use LDAP
filters in ACIs, you should verify that they target the correct entries and attributes by using the
same filter in an ldapsearch operation.
12.3.2.5. T argeting Attribute Values Using LDAP Filters
You can use access control to target specific attribute values. T his means that you can grant or deny
permissions on an attribute if that attribute's value meets the criteria defined in the ACI. An ACI that
440
Chapter 12. Managing Access Control
grants or denies access based on an attribute's value is called a value-based ACI.
For example, you might grant all users in your organization permission to modify the nsroledn attribute
in their own entry. However, you would also want to ensure that they do not give themselves certain key
roles, such as T op Level Adm inistrator. LDAP filters are used to check that the conditions on
attribute values are satisfied.
T o create a value-based ACI, you must use the targattrfilters keyword with the following syntax:
(targattrfilters="add=attr1:F1 && attr2:F2... && attrn:Fn,del=attr1:F1 && attr2:F2
... && attrn:Fn")
add represents the operation of creating an attribute.
del represents the operation of deleting an attribute.
attrx represents the target attributes.
Fx represents filters that apply only to the associated attribute.
When creating an entry, if a filter applies to an attribute in the new entry, then each instance of that
attribute must satisfy the filter. When deleting an entry, if a filter applies to an attribute in the entry, then
each instance of that attribute must also satisfy the filter.
When modifying an entry, if the operation adds an attribute, then the add filter that applies to that
attribute must be satisfied; if the operation deletes an attribute, then the delete filter that applies to that
attribute must be satisfied. If individual values of an attribute already present in the entry are replaced,
then both the add and delete filters must be satisfied.
For example, consider the following attribute filter:
(targattrfilters="add=nsroledn:(!(nsroledn=cn=superAdmin)) &&
telephoneNumber:(telephoneNumber=123*)")
T his filter can be used to allow users to add any role (nsroledn attribute) to their own entry, except the
superAdm in role. It also allows users to add a telephone number with a 123 prefix.
NOTE
You cannot create value-based ACIs from the Directory Server Console.
12.3.2.6. T argeting a Single Directory Entry
T argeting a single directory entry is not straightforward because it goes against the design philosophy
of the access control mechanism. However, it can be done in either of two ways:
By creating a bind rule that matches user input in the bind request with an attribute value stored in
the targeted entry. For more details, see Section 12.4.5, “Defining Access Based on Value Matching”.
By using the targetattr and targetfilter keywords.
You can use the targetattr keyword to specify an attribute that is only present in the entry you want
to target, and not in any of the entries below your target. For example, if you want to target
ou=people,dc=exam ple,dc=com , and there are not any organizational units (ou) defined below that
node, you could specify an ACI that contains targetattr=ou.
441
Red Hat D irectory Server 8.2 Administration Guide
A safer method is to use the targetfilter keyword and to specify explicitly an attribute value that
appears in the entry alone. For example, during the installation of the Directory Server, the following ACI
is created:
aci: (targetattr="*")(targetfilter=(o=NetscapeRoot))(version 3.0;
acl "Default anonymous access"; allow (read, search) userdn="ldap:///anyone";)
T his ACI can apply only to the o=NetscapeRoot entry.
T he risk associated with these method is that your directory tree might change in the future, and you
would have to remember to modify this ACI.
12.3.3. Defining Permissions
Permissions specify the type of access you are allowing or denying. You can either allow or deny
permission to perform specific operations in the directory. T he various operations that can be assigned
are known as rights.
T here are two parts to setting permissions:
Allowing or denying access
Assigning rights
12.3.3.1. Allowing or Denying Access
You can either explicitly allow or deny access permissions to the directory tree.
NOTE
From the Directory Server Console, you cannot explicitly deny access, only grant permissions.
12.3.3.2. Assigning Rights
Rights detail the specific operations a user can perform on directory data. You can allow or deny all
rights, or you can assign one or more of the following rights:
442
Chapter 12. Managing Access Control
T able 12.2. User Rights
Right
Description
Read
Indicates whether users can read directory data.
T his permission applies only to the search
operation.
Write
Indicates whether users can modify an entry by
adding, modifying, or deleting attributes. T his
permission applies to the modify and modrdn
operations.
Add
Indicates whether users can create an entry.
T his permission applies only to the add
operation.
Delete
Indicates whether users can delete an entry.
T his permission applies only to the delete
operation.
Search
Indicates whether users can search for the
directory data. Users must have Search and
Read rights in order to view the data returned as
part of a search result. T his permission applies
only to the search operation.
Compare
Indicates whether the users can compare data
they supply with data stored in the directory. With
compare rights, the directory returns a success
or failure message in response to an inquiry, but
the user cannot see the value of the entry or
attribute. T his permission applies only to the
compare operation.
Selfwrite
Indicates whether users can add or delete their
own DN from a group. T his right is used only for
group management.
Proxy
Indicates whether the specified DN can access
the target with the rights of another entry.
All
Indicates that the specified DN has all rights
(read, write, search, delete, com pare, and
selfwrite) to the targeted entry, excluding
proxy rights.
Rights are granted independently of one another. T his means, for example, that a user who is granted
add rights can create an entry but cannot delete it if delete rights have not been specifically granted.
T herefore, when planning the access control policy for your directory, you must ensure that you grant
rights in a way that makes sense for users. For example, it does not usually make sense to grant write
permission without granting read and search permissions.
443
Red Hat D irectory Server 8.2 Administration Guide
NOTE
T he proxy mechanism is very powerful and must be used sparingly. Proxy rights are granted
within the scope of the ACL, and there is no way to restrict who an entry that has the proxy right
can impersonate; that is, when you grant a user proxy rights, that user has the ability to proxy for
any user under the target; there is no way to restrict the proxy rights to only certain users. For
example, if an entity has proxy rights to the dc=exam ple,dc=com tree, that entity can do
anything. Make sure you set the proxy ACI at the lowest possible level of the DIT ; see
Section 12.9.12, “Proxied Authorization ACI Example”.
12.3.3.3. Rights Required for LDAP Operations
T his section describes the rights you need to grant to users depending on the type of LDAP operation
you want to authorize them to perform.
Adding an entry:
Grant add permission on the entry being added.
Grant write permission on the value of each attribute in the entry. T his right is granted by default
but could be restricted using the targattrfilters keyword.
Deleting an entry:
Grant delete permission on the entry to be deleted.
Grant write permission on the value of each attribute in the entry. T his right is granted by default
but could be restricted using the targattrfilters keyword.
Modifying an attribute in an entry:
Grant write permission on the attribute type.
Grant write permission on the value of each attribute type. T his right is granted by default but
could be restricted using the targattrfilters keyword.
Modifying the RDN of an entry:
Grant write permission on the entry.
Grant write permission on the attribute type used in the new RDN.
Grant write permission on the attribute type used in the old RDN, if you want to grant the right to
delete the old RDN.
Grant write permission on the value of attribute type used in the new RDN. T his right is granted
by default but could be restricted using the targattrfilters keyword.
Comparing the value of an attribute:
Grant compare permission on the attribute type.
Searching for entries:
Grant search permission on each attribute type used in the search filter.
Grant read permission on attribute types used in the entry.
T he permissions granted on individual attributes or entries can affect a broad range of actions; for
example, there are several different permissions users must have to search the directory like the
following ldapsearch operation:
ldapsearch -h host -s base -b "uid=bkolics,dc=example,dc=com" objectclass=* mail
T he following ACI is used to determine whether user bkolics can be granted access:
444
Chapter 12. Managing Access Control
aci: (targetattr = "mail")(version 3.0; acl "self access to
mail"; allow (read, search) userdn = "ldap:///self";)
T he search result list is empty because this ACI does not grant access to the objectclass attribute. If
you want the search operation described above to be successful, modify the ACI to allow read and
search access for the mail and objectclass attributes.
aci: (targetattr = "mail || objectclass")(version 3.0; acl "self
access to mail"; allow (read, search) userdn = "ldap:///self";)
12.3.3.4 . Permissions Syntax
In an ACI statement, the syntax for permissions is allow|deny (rights). rights is a list of 1 to 8
comma-separated keywords enclosed within parentheses. Valid keywords are read, write, add,
delete, search, com pare, selfwrite, proxy, or all.
In the following example, read, search, and compare access is allowed, provided the bind rule is
evaluated to be true:
aci: (target="ldap:///dc=example,dc=com") (version 3.0;acl "example";
allow (read, search, compare) bind_rule;)
12.3.3.5. Access Control and the modrdn Operation
T o explicitly deny m odrdn rights using ACIs, target the relevant entries but omit the targetattr
keyword. For example, to prevent the cn=helpDeskGroup,ou=groups,dc=exam ple,dc=com
group from renaming any entries in the set specified by the pattern
cn=* ,ou=people,dc=exam ple,dc=com , add the following ACI:
aci: (target="ldap:///cn=*,ou=people,dc=example,dc=com")
(version 3.0; acl "Deny modrdn rights to the helpDeskGroup";
deny(write) groupdn="ldap:///cn=helpDeskGroup,ou=groups,dc=example,dc=com";)
12.4. Bind Rules
Depending on the ACIs defined for the directory, for certain operations, you need to bind to the directory.
Binding means logging in or authenticating yourself to the directory by providing credentials (for example,
a keytab for SASL or a client certificate for SSL). T he credentials provided in the bind operation and the
circumstances of the bind determine whether access to the directory is allowed or denied.
Every permission set in an ACI has a corresponding bind rule that details the required credentials and
bind parameters.
Bind rules can be simple, such as stating that the person accessing the directory must belong to a
specific group. Bind rules can also be more complex, such as requiring that a person must belong to a
specific group, must log in from a machine with a specific IP address, and is restricted to access
between 8 a.m. and 5 p.m.
Bind rules define who can access the directory, when, and from where by defining any of the following:
Users, groups, and roles that are granted access.
Locations from which an entity must bind.
T imes or days on which binding must occur.
445
Red Hat D irectory Server 8.2 Administration Guide
T ypes of authentication that must be in use during binding.
Additionally, bind rules can be complex constructions that combine these criteria by using Boolean
operators. See Section 12.4.11, “Using Boolean Bind Rules” for more information.
12.4.1. Bind Rule Syntax
Whether access is allowed or denied depends on whether an ACI's bind rule is evaluated to be true.
Bind rules use one of the two following patterns:
keyword = "expression"; or
keyword != "expression";
Equal (=) indicates that keyword and expression must match in order for the bind rule to be true, and not
equal (!=) indicates that keyword and expression must not match in order for the bind rule to be true.
NOTE
T he tim eofday keyword also supports the inequality expressions (<, <=, >,>=). T his is the only
keyword that supports these expressions.
T he quotation marks ("") around expression and the delimiting semicolon (;) are required. T he
expressions you can use depend on the associated keyword.
T able 12.3, “LDIF Bind Rule Keywords” lists each keyword and the associated expressions and
indicates whether wildcard characters are allowed in the expression.
446
Chapter 12. Managing Access Control
T able 12.3. LDIF Bind Rule Keywords
Keyword
userdn
Valid Expressions
ldap:///distinguished_name
Wildcard Allowed
Yes, in DN only
ldap:///all
ldap:///anyone
ldap:///self
ldap:///parent
ldap:///suffix??scope?(filter)
groupdn
ldap:///DN|| DN
No
ldap:///suffix??scope?(filter)
roledn
ldap:///DN|| DN
No
userattr
attribute#bindType
orattribute#value
No
ip
IP_address
Yes
dns
DNS_host_name
Yes
dayofweek
sun mon tue wed thu fri sat
No
timeofday
0 - 2359
No
authmethod
none
No
simple
ssl
sasl sasl_mechanism
12.4.2. Defining User Access - userdn Keyword
User access is defined using the userdn keyword. T he userdn keyword requires one or more valid
distinguished names in the following format:
userdn = "ldap:///dn [|| ldap:///dn]...[||ldap:///dn]"
dn can be a DN or one of the expressions anyone, all, self, or parent:
userdn
userdn
userdn
userdn
= "ldap:///anyone" Defines anonymous access
= "ldap:///all" Defines general access
=ldap:///self" Defines self access
=ldap:///parent" Defines access for the parent entry
T he userdn keyword can also be expressed as an LDAP filter:
447
Red Hat D irectory Server 8.2 Administration Guide
ldap:///suffix??scope?(filter)
NOTE
If a DN contains a comma, the comma must be preceded by a backslash (\) escape character.
12.4 .2.1. Anonymous Access (anyone Keyword)
Granting anonymous access to the directory means that anyone can access it without providing a bind
DN or password and regardless of the circumstances of the bind. You can limit anonymous access to
specific types of access (for example, read or search access) or to specific subtrees or individual
entries within the directory.
From the Directory Server Console, you define anonymous access through the Access Control
Editor. See Section 12.5, “Creating ACIs from the Console”.
12.4 .2.2. General Access (all Keyword)
You can use bind rules to indicate that a permission applies to anyone who has successfully bound to
the directory; that is, all authenticated users. T his allows general access while preventing anonymous
access.
From the Directory Server Console, you define general access on the Access Control Editor. For
more information, see Section 12.5, “Creating ACIs from the Console”.
12.4 .2.3. Self Access (self Keyword)
Specifies that users are granted or denied access to their own entries. In this case, access is granted or
denied if the bind DN matches the DN of the targeted entry.
Self access can be configured in the Access Control Editor in the Directory Server Console. For
more information, see Section 12.5, “Creating ACIs from the Console”.
12.4 .2.4 . Parent Access (parent Keyword)
Specifies that users are granted or denied access to the entry only if their bind DN is the parent of the
targeted entry.
You cannot set up parent access control using the Directory Server Console.
12.4 .2.5. LDAP URLs
You can dynamically target users in ACIs using a URL with an LDAP filter:
userdn = "ldap:///suffix??scope?(filter)"
For example, all users in the accounting and engineering branches of the exam ple tree would be
granted or denied access to the targeted resource dynamically based on the following URL:
userdn = "ldap:///dc=example,dc=com??sub?(|(ou=engineering)(ou=accounting))"
448
Chapter 12. Managing Access Control
NOTE
Do not specify a hostname or port number within the LDAP URL. LDAP URLs always apply to the
local server.
It is possible to string multiple LDAP URLs together so that the bind user must match both filter A and
filter B. T his is done by using multiple userdn keyword definitions. For example:
userdn="ldap:///dc=example,dc=com??sub?(ou=engineering)" and
userdn="ldap:///dc=example,dc=com??sub?
(manager="uid=bjensen,ou=managers,dc=example,dc=com")"
Using a connector such as & & is not allowed. For example, this is not an acceptable bind rule:
groupdn="ldap:///dc=example,dc=com??sub?(ou=engineering) &&
ldap:///dc=example,dc=com??sub?
(manager="uid=bjensen,ou=managers,dc=example,dc=com")"
For more information about LDAP URLs, see Appendix B, LDAP URLs.
12.4 .2.6. Wildcards
You can also specify a set of users by using the wildcard character (*). For example, specifying a user
DN of uid=u* ,dc=exam ple,dc=com indicates that only users with a bind DN beginning with the
letter u are allowed or denied access based on the permissions you set.
From the Directory Server Console, you set user access from the Access Control Editor. For more
information, see Section 12.5, “Creating ACIs from the Console”.
12.4 .2.7. Examples
449
Red Hat D irectory Server 8.2 Administration Guide
T able 12.4 . userdn Keyword Examples
Example
S
c
e
n
a
r
i
o
Description
userdn
U
=
"ldap:///uid=*,dc=example,dc=com";
s
e
r
d
n
k
e
y
w
o
r
d
c
o
n
t
a
i
n
i
n
g
a
n
L
D
A
P
U
R
L
T he bind rule is evaluated to be true if the user binds to
the directory using any distinguished name of the
specified pattern. For example, both of the following bind
DNs would be evaluated to be true:
uid=ssarette,dc=example,dc=com
uid=tjaz,ou=Accounting,dc=example,dc=com
T his bind DN would be evaluated to be false:
cn=Babs Jensen,dc=example,dc=com
userdn="ldap:///uid=bj,dc=example,dc=com
U
T he bind rule is evaluated to be true if the client binds as
||s ldap:///uid=kc,dc=example,dc=com";
either of the two supplied distinguished names.
e
r
d
n
k
e
y
w
o
r
4 50
Chapter 12. Managing Access Control
r
d
c
o
n
t
a
i
n
i
n
g
l
o
g
i
c
a
l
O
R
o
f
L
D
A
P
U
R
L
s
userdn
U
!=
"ldap:///uid=*,ou=Accounting,dc=example,d
s
c=com";
e
r
d
n
k
e
y
w
o
r
d
e
x
c
l
u
d
i
n
g
a
T he bind rule is evaluated to be true if the client is not
binding as a UID-based distinguished name in the
accounting subtree. T his bind rule only makes sense if
the targeted entry is not under the accounting branch of
the directory tree.
4 51
Red Hat D irectory Server 8.2 Administration Guide
a
s
p
e
c
i
f
i
c
L
D
A
P
U
R
L
userdn
U
= "ldap:///self";
s
e
r
d
n
k
e
y
w
o
r
d
c
o
n
t
a
i
n
i
n
g
s
e
l
f
k
e
y
w
o
r
d
T he bind rule is evaluated to be true if the user is
accessing the entry represented by the DN with which the
user bound to the directory. T hat is, if the user has bound
as uid=ssarette, dc=exam ple,dc=com and the user
is attempting an operation on the
uid=ssarette,dc=exam ple,dc=com entry, then the
bind rule is true.
userdn
U
= "ldap:///all";
s
e
r
T he bind rule is evaluated to be true for any valid bind DN.
T o be true, a valid distinguished name must be presented
by the user for a successful bind operation.
4 52
If you want to grant all users in the exam ple tree write
access to their userPassword attribute, you would create
the following ACI on the dc=exam ple,dc=com node.
aci: (targetattr = "userPassword") (version 3.0; acl "writeself"; allow (write) userdn = "ldap:///self";)
Chapter 12. Managing Access Control
d
n
k
e
y
w
o
r
d
c
o
n
t
a
i
n
i
n
g
t
h
e
a
l
l
k
e
y
w
o
r
d
For example, if you want to grant read access to the entire
tree to all authenticated users, you would create the
following ACI on the dc=exam ple,dc=com node:
userdn
U
= "ldap:///anyone";
s
e
r
d
n
k
e
y
w
o
r
d
c
o
n
t
a
i
n
i
n
T he bind rule is evaluated to be true for anyone; use this
keyword to provide anonymous access to your directory.
aci:(version 3.0; acl "all-read"; allow (read)
userdn="ldap:///all";)
For example, if you want to allow anonymous read and
search access to the entire exam ple tree, you would
create the following ACI on the dc=exam ple,dc=com
node:
aci: (version 3.0; acl "anonymous-read-search"; allow
(read,search) userdn = "ldap:///anyone";)
4 53
Red Hat D irectory Server 8.2 Administration Guide
n
g
t
h
e
a
n
y
o
n
e
k
e
y
w
o
r
d
userdn
U
= "ldap:///parent";
s
e
r
d
n
k
e
y
w
o
r
d
c
o
n
t
a
i
n
i
n
g
t
h
e
p
a
r
e
n
t
k
e
y
w
4 54
T he bind rule is evaluated to be true if the bind DN is the
parent of the targeted entry.
For example, if you want to grant write access to every
user's child entries, you would create the following ACI on
the dc=exam ple,dc=com node:
aci:(version 3.0; acl "parent access"; allow (write)
userdn="ldap:///parent";)
Chapter 12. Managing Access Control
o
r
d
12.4.3. Defining Group Access - groupdn Keyword
Members of a specific group can access a targeted resource. T his is known as group access. Group
access is defined using the groupdn keyword to specify that access to a targeted entry is granted or
denied if the user binds using a DN that belongs to a specific group.
Group membership can be determined based on the user's DN or by using an LDAP filter to search for
group members.
T he groupdn keyword requires one or more valid distinguished names in the following format:
groupdn="ldap:///dn [|| ldap:///dn]...[|| ldap:///dn]"
T he bind rule is evaluated to be true if the bind DN belongs to the named group.
NOTE
If a DN contains a comma, the comma must be escaped by a backslash (\).
T he groupdn keyword can also be expressed with an LDAP filter:
groupdn="ldap:///suffix??scope?(filter)
With more complex groupdn syntax, the value of the groupdn expression is a single LDAP URL.
Multiple groupdns can be grouped together within parentheses and use or or and connectors to
define additional conditions on the group membership. For example:
(groupdn = "ldap:///ou=Groups,dc=example,dc=com??sub?(cn=*s_0)" or groupdn =
"ldap:///ou=Groups,dc=example,dc=com??sub?(cn=*s_1)") and groupdn =
"ldap:///ou=Groups,dc=example,dc=com??sub?(cn=*s_2)"
When stringing multiple groupdn URLs together, the keyword supports pipes to separate the URLs:
groupdn = "LDAPURI0 || LDAPURL1 || LDAPURL2"
However, it is not permissible to use ampersands (& ), like groupdn = "LDAPURI0 & & LDAPURL1",
or double quotes.
For example, to use two groupdn keywords so that the bind user must belong to both an Administrators
group and a Managers group:
groupdn="ldap:///dc=example,dc=com??sub?(cn=*Administrators)" and
groupdn="ldap:///dc=example,dc=com??sub?(cn=*Managers)"
For more information about LDAP URLs, see Appendix B, LDAP URLs.
T he Directory Server Console defines specific groups through the Access Control Editor. For
4 55
Red Hat D irectory Server 8.2 Administration Guide
more information, see Section 12.5, “Creating ACIs from the Console”.
4 56
Chapter 12. Managing Access Control
T able 12.5. groupdn Examples
Example
S
c
e
n
a
r
i
o
Description
groupdn
G
=
"ldap:///cn=Administrators,dc=example,dc=
r
com";
o
u
p
d
n
k
e
y
w
o
r
d
c
o
n
t
a
i
n
i
n
g
a
n
L
D
A
P
U
R
L
T he bind rule is evaluated to be true if the bind DN
belongs to the Administrators group. If you wanted to
grant the Administrators group permission to write to the
entire directory tree, you would create the following ACI on
the dc=exam ple,dc=com node:
groupdn
G
= "ldap:///dc=example,dc=com??
sub?(cn=*Administrators)";
r
o
u
p
d
n
k
e
y
w
T he bind rule is evaluated to be true if the bind DN
belongs to any of the groups which are returned, meaning
they match the filter.
aci: (targetattr=*)(version 3.0; acl "Administrators-write";
allow (write)
groupdn="ldap:///cn=Administrators,dc=example,dc=com";
)
4 57
Red Hat D irectory Server 8.2 Administration Guide
w
o
r
d
c
o
n
t
a
i
n
i
n
g
a
n
L
D
A
P
U
R
L
w
i
t
h
a
f
i
l
t
e
r
groupdn
G
=
"ldap:///cn=Administrators,dc=example,dc=
r
com"
o
|| "ldap:///cn=Mail
Administrators,dc=example,dc=com";
u
p
d
n
k
e
y
w
o
r
d
c
o
n
t
a
i
4 58
T he bind rule is evaluated to be true if the bind DN
belongs to either the Adm inistrators or the Mail
Adm inistrators group.
Chapter 12. Managing Access Control
i
n
i
n
g
l
o
g
i
c
a
l
O
R
o
f
L
D
A
P
U
R
L
s
12.4.4. Defining Role Access - roledn Keyword
Members of a specific role can access a targeted resource. T his is known as role access. Role access
is defined using the roledn keyword to specify that access to a targeted entry is granted or denied if
the user binds using a DN that belongs to a specific role.
T he roledn keyword requires one or more valid distinguished names in the following format:
roledn = "ldap:///dn [|| ldap:///dn]... [|| ldap:///dn]"
T he bind rule is evaluated to be true if the bind DN belongs to the specified role.
NOTE
If a DN contains a comma, the comma must be escaped by a backslash (\).
T he roledn keyword has the same syntax and is used in the same way as the groupdn keyword, with
the exception of the LDAP filter, which is not implemented for role membership.
12.4.5. Defining Access Based on Value Matching
You can set bind rules to specify that an attribute value of the entry used to bind to the directory must
match an attribute value of the targeted entry.
For example, you can specify that the bind DN must match the DN in the manager attribute of a user
entry in order for the ACI to apply. In this case, only the user's manager would have access to the entry.
T his example is based on DN matching. However, you can match any attribute of the entry used in the
4 59
Red Hat D irectory Server 8.2 Administration Guide
bind with the targeted entry. For example, you could create an ACI that allowed any user whose
favoriteDrink attribute is beer to read all the entries of other users that have the same value for
favoriteDrink.
12.4 .5.1. Using the userattr Keyword
T he userattr keyword can be used to specify which attribute values must match between the entry
used to bind and the targeted entry. You can specify any of the following:
A user DN
A group DN
A role DN
An LDAP filter, in an LDAP URL
Any attribute type
T he LDIF syntax of the userattr keyword is as follows:
userattr = "attrName#bindType
Using an attribute type that requires a value other than a user DN, group DN, role DN, or an LDAP filter
has the following format:
userattr = "attrName#attrValue
attrName is the name of the attribute used for value matching.
bindType is either USERDN, GROUPDN, or LDAPURL.
attrValue is any string representing an attribute value.
12.4 .5.1.1. Example with USERDN Bind T ype
T he following associates the userattr keyword with a bind based on the user DN:
userattr = "manager#USERDN"
T he bind rule is evaluated to be true if the bind DN matches the value of the manager attribute in the
targeted entry. You can use this to allow a user's manager to modify employees' attributes. T his
mechanism only works if the manager attribute in the targeted entry is expressed as a full DN.
T he following example grants a manager full access to his or her employees' entries:
aci: (target="ldap:///dc=example,dc=com")(targetattr=*)
(version 3.0; acl "manager-write"; allow (all) userattr = "manager#USERDN";)
12.4 .5.1.2. Example with GROUPDN Bind T ype
T he following associates the userattr keyword with a bind based on a group DN:
userattr = "owner#GROUPDN"
T he bind rule is evaluated to be true if the bind DN is a member of the group specified in the owner
attribute of the targeted entry. For example, you can use this mechanism to allow a group to manage
employees' status information. You can use an attribute other than owner as long as the attribute you
use contains the DN of a group entry.
4 60
Chapter 12. Managing Access Control
T he group you point to can be a dynamic group, and the DN of the group can be under any suffix in the
database. However, the evaluation of this type of ACI by the server is very resource intensive.
If you are using static groups that are under the same suffix as the targeted entry, you can use the
following expression:
userattr = "ldap:///dc=example,dc=com?owner#GROUPDN"
In this example, the group entry is under the dc=exam ple,dc=com suffix. T he server can process this
type of syntax more quickly than the previous example.
(By default, owner is not an allowed entry in a user's entry. You would have to extend your schema to
allow this attribute in a person object.)
12.4 .5.1.3. Example with ROLEDN Bind T ype
T he following associates the userattr keyword with a bind based on a role DN:
userattr = "exampleEmployeeReportsTo#ROLEDN"
T he bind rule is evaluated to be true if the bind DN belongs to the role specified in the
exampleEmployeeReportsTo attribute of the targeted entry. For example, if you create a nested role for
all managers in your company, you can use this mechanism to grant managers at all levels access to
information about employees that are at a lower grade than themselves.
NOTE
T his example assumes that you have added the exampleEmployeeReportsToattribute to the
schema and that all employee entries contain this attribute. It also assumes that the value of this
attribute is the DN of a role entry. For information on adding attributes to the schema, see
Section 6.4.2, “Creating Attributes”.
T he DN of the role can be under any suffix in the database. If you are also using filtered roles, the
evaluation of this type of ACI uses a lot of resources on the server.
If you are using a static role definition and the role entry is under the same suffix as the targeted entry,
you can use the following expression:
userattr = "ldap:///dc=example,dc=com?employeeReportsTo#ROLEDN"
In this example, the role entry is under the dc=exam ple,dc=com suffix. T he server can process this
type of syntax more quickly than the previous example.
12.4 .5.1.4 . Example with LDAPURL Bind T ype
T he following associates the userattr keyword with a bind based on an LDAP filter:
userattr = "myfilter#LDAPURL
T he bind rule is evaluated to be true if the bind DN matches the filter specified in the myfilter attribute of
the targeted entry. T he myfilter attribute can be replaced by any attribute that contains an LDAP filter.
12.4 .5.1.5. Example with Any Attribute Value
4 61
Red Hat D irectory Server 8.2 Administration Guide
T he following associates the userattr keyword with a bind based on any attribute value:
userattr = "favoriteDrink#Beer"
T he bind rule is evaluated to be true if the bind DN and the target DN include the favoriteDrink
attribute with a value of Beer.
12.4 .5.1.6. Using the userattr Keyword with Inheritance
When you use the userattr keyword to associate the entry used to bind with the target entry, the ACI
applies only to the target specified and not to the entries below it. In some circumstances, you might
want to extend the application of the ACI several levels below the targeted entry. T his is possible by
using the parent keyword and specifying the number of levels below the target that should inherit the
ACI.
When you use the userattr keyword in association with the parent keyword, the syntax is as follows:
userattr = "parent[inheritance_level].attrName#bindType
Using an attribute type that requires a value other than a user DN, group DN, role DN, or an LDAP filter,
the syntax is as follows:
userattr = "parent[inheritance_level].attrName#attrValue
inheritance_level is a comma-separated list that indicates how many levels below the target inherits
the ACI. You can include five levels (0, 1, 2, 3, 4 ) below the targeted entry; zero (0) indicates the
targeted entry.
attribute is the attribute targeted by the userattr or groupattr keyword.
bindType can be one of USERDN, GROUPDN, or LDAPURL.
For example:
userattr = "parent[0,1].manager#USERDN"
T his bind rule is evaluated to be true if the bind DN matches the manager attribute of the targeted entry.
T he permissions granted when the bind rule is evaluated to be true apply to the target entry and to all
entries immediately below it.
T he example in Figure 12.1, “Using Inheritance With the userattr Keyword” indicates that user bjensen
is allowed to read and search the cn=Profiles entry as well as the first level of child entries which
includes cn=m ail and cn=news, thus allowing her to search through her own mail and news IDs.
4 62
Chapter 12. Managing Access Control
Figure 12.1. Using Inheritance With the userattr Keyword
In this example, if you did not use inheritance, you would have to do one of the following to achieve the
same result:
Explicitly set read and search access for user bjensen on the cn=Profiles, cn=m ail, and
cn=news entries in the directory.
Add the owner attribute with a value of bjensen to the cn=m ail and cn=news entries, and then
add the following ACI to the cn=m ail and cn=news entries.
aci: (targetattr="*") (version 3.0; acl "profiles access"; allow (read,search)
userattr="owner#USERDN";)
12.4 .5.1.7. Granting Add Permission Using the userattr Keyword
Using the userattr keyword in conjunction with all or add permissions does not behave as one
would typically expect. T ypically, when a new entry is created in the directory, Directory Server evaluates
access rights on the entry being created and not on the parent entry. However, in the case of ACIs using
the userattr keyword, this behavior could create a security hole, and the server's normal behavior is
modified to avoid it.
Consider the following example:
aci: (target="ldap:///dc=example,dc=com")(targetattr=*) (version 3.0;
acl "manager-write"; allow (all) userattr = "manager#USERDN";)
T his ACI grants managers all rights on the entries of employees that report to them. However, because
access rights are evaluated on the entry being created, this type of ACI would also allow any employee
to create an entry in which the manager attribute is set to their own DN. For example, disgruntled
employee Joe (cn=Joe,ou=eng,dc=exam ple,dc=com ) might want to create an entry in the Human
Resources branch of the tree to use (or misuse) the privileges granted to Human Resources
employees.
He could do this by creating the following entry:
dn: cn= Trojan Horse,ou=Human Resources,dc=example,dc=com
objectclass: top
...
cn: Trojan Horse
manager: cn=Joe,ou=eng,dc=example,dc=com
4 63
Red Hat D irectory Server 8.2 Administration Guide
T o avoid this type of security threat, the ACI evaluation process does not grant add permission at level
0, to the entry itself. You can, however, use the parent keyword to grant add rights below existing
entries. You must specify the number of levels below the parent for add rights. For example, the following
ACI allows child entries to be added to any entry in the dc=exam ple,dc=com that has a manager
attribute that matches the bind DN:
aci: (target="ldap:///dc=example,dc=com")(targetattr=*)
(version 3.0; acl "parent-access"; allow (add)
userattr = "parent[0,1].manager#USERDN";)
T his ACI ensures that add permission is granted only to users whose bind DN matches the manager
attribute of the parent entry.
12.4.6. Defining Access from a Specific IP Address
Using bind rules, you can indicate that the bind operation must originate from a specific IP address. T his
is often used to force all directory updates to occur from a given machine or network domain.
T he LDIF syntax for setting a bind rule based on an IP address is as follows:
ip = "IP_address" or ip != "IP_address"
T he IP address must be expressed in dot notation. You can use the wildcard character (*) to include
multiple machines. For example, the following string is valid:
ip = "12.123.1.*";
T he bind rule is evaluated to be true if the client accessing the directory is located at the named IP
address. T his can be useful for allowing certain kinds of directory access only from a specific subnet or
machine.
For example, use a wildcard IP address such as 12.3.4 5.* to specify a specific subnetwork or
123.4 5.6.* +255.255.255.115 to specify a subnetwork mask.
From the Directory Server Console, you can define specific machines to which the ACI applies through
the Access Control Editor. For more information, see Section 12.5, “Creating ACIs from the
Console”.
12.4.7. Defining Access from a Specific Domain
A bind rule can specify that the bind operation must originate from a particular domain or host machine.
T his is often used to force all directory updates to occur from a given machine or network domain.
T he LDIF syntax for setting a bind rule based on the DNS hostname is as follows:
dns = "DNS_Hostname or dns != "DNS_Hostname
WARNING
T he dns keyword requires that the naming service used on your machine is DNS. If the name
service is not DNS, use the ip keyword instead.
4 64
Chapter 12. Managing Access Control
T he dns keyword requires a fully qualified DNS domain name. Granting access to a host without
specifying the domain creates a potential security threat. For example, the following expression is
allowed but not recommended:
dns = "legend.eng";
Instead, use a fully qualified name:
dns = "legend.eng.example.com";
T he dns keyword allows wildcards. For example:
dns = "*.example.com";
T he bind rule is evaluated to be true if the client accessing the directory is located in the named domain.
T his can be useful for allowing access only from a specific domain. Wildcards will not work if your
system uses a naming service other than DNS. In such a case, if you want to restrict access to a
particular domain, use the ip keyword, as described in Section 12.4.6, “Defining Access from a Specific
IP Address”.
12.4.8. Requiring a Certain Level of Security in Connections
A bind rule can specify that an operation must occur over a connection with a certain level of security.
T his forces operations — such as password change operations — to be performed over an SSL/T LS,
Start T LS, or SASL connection. T he secureness of the connection is determined by its security strength
factor, which sets the minimum key strength required to process operations. T he value for the SSF for
any operation is the higher of the values between an SSL/T LS connection and a SASL bind; this means
that if a server is configured to run over T LS and a replication agreement is configured for SASL/GSSAPI,
the SSF for the operation is whichever available encryption type is more secure.
T he SSF can be specified using any comparison pattern, like equals (=), less-than and greater-than, (<
and >), is-not (!), less-than-or-equal-to (<=), or greater-than-or-equal-to (>=). T he LDIF syntax for
setting a bind rule based on the SSF is as follows:
ssf = "key_strength"
ssf >= "key_strength"
T he ssf keyword accepts any positive whole number. If this is set to 0, than no secure connection is
required for an operation.
T he bind rule is evaluated to be true if the client accesses the directory using a connection of adequate
strength.
12.4.9. Defining Access at a Specific Time of Day or Day of Week
You can use bind rules to specify that binding can only occur at a certain time of day or on a certain day
of the week. For example, you can set a rule that allows access only if it is between the hours of 8 a.m.
and 5 p.m. Monday through Friday. T he time used to evaluate access rights is the time on the Directory
Server, not the time on the client.
T he LDIF syntax for setting a bind rule based on the time of day is as follows:
timeofday operator time
4 65
Red Hat D irectory Server 8.2 Administration Guide
operator can be one of the following symbols:
equal to (=)
not equal to (!=)
greater than (>)
greater than or equal to (>=)
less than (<)
less than or equal to (<=)
T he tim eofday keyword requires a time of day expressed in hours and minutes in the 24 hour clock
(0 to 2359).
NOTE
T he time on the Directory Server is used for the evaluation, not the time on the client.
T he LDIF syntax for setting a bind rule based on the day in the week is as follows:
dayofweek = "day1, day2 ...
T he possible values for the dayofweek keyword are the English three-letter abbreviations for the days
of the week: sun, m on, tue, wed, thu, fri, sat.
12.4 .9.1. Examples
T he following are examples of the tim eofday and dayofweek syntax:
T he bind rule is evaluated to be true if the client is accessing the directory at noon.
timeofday = "1200";
T he bind rule is evaluated to be true if the client is accessing the directory at any time other than 1
a.m.
timeofday != "0100";
T he bind rule is evaluated to be true if the client is accessing the directory at any time after 8 a.m.
timeofday > "0800";
T he bind rule is evaluated to be true if the client is accessing the directory at any time before 6 p.m.
timeofday < "1800";
T he bind rule is evaluated to be true if the client is accessing the directory at 8 a.m. or later.
timeofday >= "0800";
T he bind rule is evaluated to be true if the client is accessing the directory at 6 p.m. or earlier.
timeofday <= "1800";
4 66
Chapter 12. Managing Access Control
T he bind rule is evaluated to be true if the client is accessing the directory on Sunday, Monday, or
T uesday.
dayofweek = "Sun, Mon, Tue";
12.4.10. Defining Access Based on Authentication Method
T he authm ethod keyword sets the specific method that a client uses to bind to the directory. T here are
four available authentication methods:
None. Authentication is not required. T his is the default. It represents anonymous access.
Simple. T he client must provide a user name and password to bind to the directory.
SSL. T he client must bind to the directory using some kind of PKI credentials, meaning a client must
present an SSL certificate either in a database or on a smart card, token, or some other device.
Certificate-based authentication, as one method, is described in Section 14.2.10, “Using CertificateBased Authentication”.
SASL. T he client must bind to the directory over a Simple Authentication and Security Layer (SASL)
connection. Directory Server supports three SASL mechanisms: EXT ERNAL, CRAM-MD5, DIGEST MD5 (for Kerberos systems), and GSS-API (for Kerberos systems). For information on setting up
SASL, see Section 14.3, “Using SASL”.
NOTE
You cannot set up authentication-based bind rules through the Access Control Editor.
T he LDIF syntax for setting a bind rule based on an authentication method is as follows:
authmethod = "sasl_mechanism
sasl_mechanism can be none, sim ple, ssl, or "sasl sasl_mechanism".
12.4 .10.1. Examples
T he following are examples of the authm ethod keyword:
Authentication is not checked during bind rule evaluation.
authmethod = "none";
T he bind rule is evaluated to be true if the client is accessing the directory using a username and
password.
authmethod = "simple";
T he bind rule is evaluated to be true if the client authenticates to the directory using a certificate over
LDAPS. T his is not evaluated to be true if the client authenticates using simple authentication (bind
DN and password) over LDAPS. T he authm ethod = "ssl" means that a certificate must be
presented to authenticate to the server. T his does not configure a required connection type, even
though SSL has to be used with certificate-based authentication.
authmethod = "ssl";
4 67
Red Hat D irectory Server 8.2 Administration Guide
T he bind rule is evaluated to be true if the client is accessing the directory using the SASL DIGEST MD5 mechanism.
authmethod = "sasl DIGEST-MD5";
12.4.11. Using Boolean Bind Rules
Bind rules can be complex expressions that use the Boolean expressions AND, OR, and NOT to set very
precise access rules. You cannot use the Directory Server Console to create Boolean bind rules. You
must create an LDIF statement.
T he LDIF syntax for a Boolean bind rule is as follows:
bind_rule [boolean][bind_rule][boolean][bind_rule]...;)
For example, this bind rule is evaluated to be true if the bind DN is a member of either the administrator's
group or the Mail Adm inistrator's group and if the client is running from within the exam ple.com
domain:
(groupdn = "ldap:///cn=administrators,dc=example,dc=com" or
groupdn = "ldap:///cn=mail administrators,dc=example,dc=com" and
dns = "*.example.com";)
T he trailing semicolon (;) is a required delimiter that must appear after the final bind rule.
Boolean expressions are evaluated in the following order:
Innermost to outermost parenthetical expressions first.
All expressions from left to right.
NOT before AND or OR operators.
OR and AND operators have no order of precedence.
Consider the following Boolean bind rules:
(bind_rule_A) OR (bind_rule_B)
(bind_rule_B) OR (bind_rule_A)
Because Boolean expressions are evaluated from left to right, in the first case, bind rule A is evaluated
before bind rule B, and, in the second case, bind rule B is evaluated before bind rule A.
However, the Boolean NOT is evaluated before the Boolean OR and Boolean AND. T hus, in the following
example, bind rule B is evaluated before bind rule A despite the left-to-right rule.
(bind_rule_A) AND NOT (bind_rule_B)
12.5. Creating ACIs from the Console
You can use the Directory Server Console to view, create, edit, and delete access control instructions for
your directory:
Section 12.5.1, “Displaying the Access Control Editor”
Section 12.5.2, “Creating a New ACI”
Section 12.5.3, “Editing an ACI”
4 68
Chapter 12. Managing Access Control
Section 12.5.4, “Deleting an ACI”
See Section 12.9, “Access Control Usage Examples” for a collection of access control rules commonly
used in Directory Server security policies, along with step-by-step instructions for using the Directory
Server Console to create them.
T he Access Control Editor prevents creating more complex ACIs in visual editing mode, especially
ACIs with any of these characteristics:
Deny access (Section 12.3.3.4, “Permissions Syntax”).
Create value-based ACIs (Section 12.3.2.2, “T argeting Attributes”).
Define parent access (Section 12.4.2.4, “Parent Access (parent Keyword)”).
Create ACIs that contain Boolean bind rules (Section 12.4.11, “Using Boolean Bind Rules”).
Create ACIs that use the roledn, userattr, authm ethod keywords.
NOTE
In the Access Control Editor, click the Edit Manually button at any time to check the
LDIF representation of the ACI changes made through the graphical interface.
12.5.1. Displaying the Access Control Editor
1. Start the Directory Server Console. Log in using the bind DN and password of a privileged user,
such as the Directory Manager, who has write access to the ACIs configured for the directory.
2. Select the Directory tab.
3. Right-click the entry in the navigation tree for which to set access control, and select Set Access
Perm issions from the pop-up menu.
Alternatively, highlight the entry, and select Set Access Perm issions from the Object menu.
4. Click New to open the Access Control Editor.
4 69
Red Hat D irectory Server 8.2 Administration Guide
Figure 12.2. Access Control Editor Window
12.5.2. Creating a New ACI
T o create a new ACI in the Directory Server Console:
1. Open the Access Control Editor, as described in Section 12.5.1, “Displaying the Access
Control Editor”.
If the view displayed is different from Figure 12.2, “Access Control Editor Window”, click the Edit
Visually button.
2. T ype the ACI name in the ACI Nam e field.
T he name can be any unique string to identify the ACI. If you do not enter a name, the server uses
unnam ed ACI.
3. In the Users/Groups tab, select the users to whom you are granting access by highlighting All
Users or clicking the Add button to search the directory for the users to add.
4 70
Chapter 12. Managing Access Control
a. Select a search area from the drop-down list, enter a search string in the Search field, and
click the Search button. You can use wildcards (an asterisk, * ) to search for partial
usernames. T he search results are displayed in the list below.
b. Highlight the entries you want in the search result list, and click the Add button to add them
to the list of entries which have access permission.
c. Click OK to dismiss the Add Users and Groups window.
T he selected entries are now listed on the Users/Groups tab in the ACI editor.
4. In the Access Control Editor, click the Rights tab, and use the checkboxes to select the
rights to grant.
4 71
Red Hat D irectory Server 8.2 Administration Guide
5. Click the T argets tab. Click T his Entry to display the current node as the target for the ACI or
click Browse to select a different suffix.
4 72
Chapter 12. Managing Access Control
NOTE
You can change the value of the target DN, but the new DN must be a direct or indirect
child of the selected entry.
If you do not want every entry in the subtree under this node to be targeted by the ACI, enter a
filter in the Filter for Sub-entries field. T he filter applies to every entry below the target
entry; for example, setting a filter of ou=Sales means that only entries with ou=Sales in their DN
are returned.
Additionally, you can restrict the scope of the ACI to only certain attributes by selecting the
attributes to target in the attribute list.
6. Click the Hosts tab, then the Add button to open the Add Host Filter dialog box.
You can specify a hostname or an IP address. With an IP address, you can use an asterisk (* ) as
a wildcard.
7. Click the T im es tab to display the table showing at what times access is allowed.
By default, access is allowed at all times. You can change the access times by clicking and
dragging the cursor over the table. You cannot select discrete blocks of time, only continuous time
ranges.
4 73
Red Hat D irectory Server 8.2 Administration Guide
8. Click OK when all of the configuration is complete.
T he Access Control Editor closes, and the new ACI is listed in the Access Control Manager
window.
NOTE
For any point of creating the ACI, click the Edit Manually button to display the LDIF statement
corresponding to the wizard input. T his statement can be edited directly, but the changes may not
be visible in the graphical interface.
12.5.3. Editing an ACI
1. In the Directory tab, right-click the top entry in the subtree, and choose Set Access
Perm issions from the pop-up menu.
4 74
Chapter 12. Managing Access Control
T he Access Control Manager window opens, listing the ACIs belonging to the entry.
2. In the Access Control Manager window, highlight the ACI to edit, and click Edit.
3. Make the edits to the ACI in the Access Control Editor; the different screens are described
more in Section 12.5.2, “Creating a New ACI” and in the online help.
4. When the edits are complete, click OK.
12.5.4. Deleting an ACI
1. In the Directory tab, right-click the top entry in the subtree, and choose Set Access
Perm issions from the pop-up menu.
4 75
Red Hat D irectory Server 8.2 Administration Guide
2. In the Access Control Manager window, select the ACI to delete.
3. Click Rem ove.
T he ACI is no longer listed in the Access Control Manager window.
12.6. Viewing ACIs
All the ACIs under a single suffix in the directory can be viewed from the command line by using the
following ldapsearch command:
[7]
/usr/lib64/mozldap/ldapsearch -h host -p port -b baseDN -D rootDN -w
rootPassword (aci=*) aci
See Chapter 8, Finding Directory Entries for information on using the ldapsearch utility.
From the Directory Server Console, all of the ACIs that apply to a particular entry can be viewed through
the Access Control Manager.
1. Start the Directory Server Console.
2. In the Directory tab, right-click the entry in the navigation tree, and select Set Access
Perm issions.
3. Check the Show Inherited ACIs checkbox to display all ACIs created on entries above the
4 76
Chapter 12. Managing Access Control
selected entry that also apply.
12.7. Checking Access Rights on Entries (Get Effective Rights)
Finding the access rights that a user has on attributes within a specific entry offers a convenient way for
administrators to find and control the access rights.
Get effective rights is a way to extend directory searches to display what access rights — such as read,
search, write and self-write, add, and delete — a user has to a specified entry.
In Directory Server, regular users can check their rights over entries which they can view and can check
other people's access to their personal entries. T he Directory Manager can check rights that one user
has over another user.
T here are two common situations where checking the effective rights on an entry are useful:
An administrator can use the get effective rights command in order to better organize access control
instructions for the directory. It is frequently necessary to restrict what one group of users can view
or edit versus another group. For instance, members of the QA Managers group may have the right
to search and read attributes like manager and salary but only HR Group members have the rights
to modify or delete them. Checking the effective rights for a user or group is one way to verify that the
appropriate access controls are in place.
A user can run the get effective rights command to see what attributes he can view or modify on his
personal entry. For instance, a user should have access to attributes such as homePostalAddress
and cn but may only have read access to manager and salary attributes.
T here are three people involved in a get effective rights search. T he first is the person running the
search command, the requester. T he rights are checked (with a variety of permutations) to see what
rights Person A has over Entry B. T he person whose rights are being checked (Person A) is the GER
subject; as in, their rights are the subject of the search. T he entry or entries to which the person has
rights (Entry B) is the target of the search or the search base.
12.7.1. Rights Shown with a Get Effective Rights Search
Any get effective rights search, both when viewing an entry in the Directory Server Console and
4 77
Red Hat D irectory Server 8.2 Administration Guide
searching for it in the command line, shows the rights that User A has to User B's entry.
T here are two kinds of access rights that can be allowed to any entry. T he first are upper-level rights,
rights on the entry itself, which means that kinds of operations that the User A can perform on User B's
entry as a whole. T he second level of access rights are more granular, show what rights for a given
attribute User A has. In this case, User A may have different kinds of access permissions for different
attributes in the same entry. Whatever access controls are allowed for a user are the effective rights
over that entry.
For example:
entryLevelRights: vadn
attributeLevelRights: givenName:rscWO, sn:rscW, objectClass:rsc, uid:rsc, cn:rscW
T able 12.6, “Entry Rights” and T able 12.7, “Attribute Rights” show the access rights to entries and
attributes, respectively, that are returned by a get effective rights search.
T able 12.6. Entry Rights
Permission
Description
a
Add an entry.
d
Delete this entry.
n
Rename the DN.
v
View the entry.
T able 12.7. Attribute Rights
Permission
Description
r
Read.
s
Search.
w
Write (m od-add).
o
Obliterate(m od-del). Analogous to delete.
c
Compare.
W
Self-write.
O
Self-delete.
12.7.2. The Format of a Get Effective Rights Search
Get effective rights (sometimes called GER) is an extended directory search; the GER parameters are
defined with the -J option with the ldapsearch command. (If an ldapsearch is run without the -J
option, then, naturally, the entry is returned as normal, without any get effective rights information.)
/usr/lib64/mozldap/ldapsearch -p port -h host -D bindDN -w bindPassword -b
searchBase -J 1.3.6.1.4.1.42.2.27.9.5.2:criticality:dn:GER_subject (searchFilter)
attributeList
-b searchBase is the base DN of the subtree or entry used to search for the GER subject.
If the search base is a specific entry DN or if only one entry is returned, then the results show the
rights the requester has over that specific entry. If multiple entries beneath the search base match
the filter, then the search returns every matching entry, with the rights for the requester over each
4 78
Chapter 12. Managing Access Control
entry.
1.3.6.1.4 .1.4 2.2.27.9.5.2 is the OID for the get effective rights control.
criticality specifies whether the search operation should return an error if the server does not
support this control (true) or if it should be ignored and let the search return as normal (false).
T he GER_subject is the person whose rights are being checked. If the GER_subject is left blank
(dn:), than the rights of an anonymous user are returned.
An optional attributeList limits the get effective rights results to the specified attribute or object class.
As with a regular ldapsearch, this can give specific attributes, like mail. If no attributes are listed,
then every present attribute for the entry is returned. Using an asterisk (* ) returns the rights for
every possible attribute for the entry, both existing attribute and non-existent attributes. Using an
plus sign (+) returns operational attributes for the entry. Examples for checking rights for specific
attributes are given in Section 12.7.2.2, “Examples of Get Effective Rights Searches for Non-Existent
Attributes” and Section 12.7.2.3, “Examples of Get Effective Rights Searches for Specific Attributes or
Object Classes”.
T he crux of a get effective rights search is the ability to check what rights the GER subject (-J) has to
the targets of the search (-b). T he get effective rights search is a regular ldapsearch, in that it simply
looks for entries that match the search parameters and returns their information. T he get effective rights
option adds extra information to those search results, showing what rights a specific user has over
those results. T hat GER subject user can be the requester himself (-D is the same as -J) or someone
else.
If the requester is a regular user (not the Directory Manager), then the requester can only see the
effective that a GER subject has on the requester's own entry. T hat is, if John Smith runs a request to
see what effective rights Babs Jensen has, then he can only get the effective rights that Babs Jensen
has on his own entry. All of the other entries return an insufficient access error for the effective rights.
T here are three general scenarios for a regular user when running a get effective rights search:
User A checks the rights that he has over other directory entries.
User A checks the rights that he has to his personal entry.
User A checks the rights that User B has to User A's entry.
T he get effective rights search has a number of flexible different ways that it can check rights on
attributes.
NOTE
All of the searches in these examples assume that the ldapsearch command is run on Red Hat
Enterprise Linux 5 (64-bit) using the Mozilla LDAP tools.
12.7.2.1. General Examples on Checking Access Rights
One common scenario for effective rights searches is for a regular user to determine what changes he
can make to his personal entry.
For example, T ed Morris wants to check the rights he has to his entry. Both the -D and -J options give
his entry as the requester. Since he is checking his personal entry, the -b option also contains his DN.
4 79
Red Hat D irectory Server 8.2 Administration Guide
Example 12.1. Checking Personal Rights (User A to User A)
/usr/lib64/mozldap/ldapsearch -p 389 -h localhost -D
"uid=tmorris,ou=people,dc=example,dc=com" -w secret -b
"uid=tmorris,ou=people,dc=example,dc=com" -J
"1.3.6.1.4.1.42.2.27.9.5.2:true:dn:uid=tmorris,ou=people,dc=example,dc=com"
"(objectClass=*)"
dn: uid=tmorris,ou=People,dc=example,dc=com
givenName: Ted
sn: Morris
ou: IT
ou: People
l: Santa Clara
manager: uid=jsmith,ou=People,dc=example,dc=com
roomNumber: 4117
mail: [email protected]
facsimileTelephoneNumber: +1 408 555 5409
objectClass: top
objectClass: person
objectClass: organizationalPerson
objectClass: inetOrgPerson
uid: tmorris
cn: Ted Morris
userPassword: {SSHA}bz0uCmHZM5b357zwrCUCJs1IOHtMD6yqPyhxBA==
entryLevelRights: v
attributeLevelRights: givenName:rsc, sn:rsc, ou:rsc, l:rsc, manager:rsc,
roomNumber:rscwo, mail:rscwo, facsimileTelephoneNumber:rscwo, objectClass:rsc,
uid:rsc, cn:rsc, userPassword:wo
T ed Morris may, for example, be a manager or work in a department where he has to edit other user's
entries, such as IT or human resources. In this case, he may want to check what rights he has to
another user's entry, as in Example 12.2, “Personally Checking the Rights of One User over Another
(User A to User B)”, where T ed (-D) checks his rights (-J) to Dave Miller's entry (-b):
Example 12.2. Personally Checking the Rights of One User over Another (User A to User
B)
/usr/lib64/mozldap/ldapsearch -p 389 -h localhost -D
"uid=tmorris,ou=people,dc=example,dc=com" -w secret -b
"uid=dmiller,ou=people,dc=example,dc=com" -J
"1.3.6.1.4.1.42.2.27.9.5.2:true:dn:uid=tmorris,ou=people,dc=example,dc=com"
"(objectClass=*)"
dn: uid=dmiller,ou=People,dc=example,dc=com
... snip ...
entryLevelRights: vad
attributeLevelRights: givenName:rscwo, sn:rscwo, ou:rscwo, l:rscwo, manager:rsc,
roomNumber:rscwo, mail:rscwo, facsimileTelephoneNumber:rscwo,
objectClass:rscwo, uid:rscwo, cn:rscwo, userPassword:rswo
For all attributes, T ed Morris has read, search, compare, modify, and delete permissions to Dave Miller's
entry. T hese results are different than the ones returned in checking T ed Morris's access to his own
entry, since he personally had only read, search, and compare rights to most of these attributes.
4 80
Chapter 12. Managing Access Control
T he Directory Manager has the ability to check the rights that one user has over another user's entry. In
Example 12.3, “T he Directory Manager's Checking the Rights of One User over Another (User A to User
B)”, the Directory Manager is checking the rights that a manager, Jane Smith (-J), has over her
subordinate, T ed Morris (-b):
Example 12.3. T he Directory Manager's Checking the Rights of One User over Another
(User A to User B)
/usr/lib64/mozldap/ldapsearch -p 389 -h localhost -D "cn=directory manager" -w
secret -b "uid=tmorris,ou=people,dc=example,dc=com" -J
"1.3.6.1.4.1.42.2.27.9.5.2:true:dn:uid=jsmith,ou=people,dc=example,dc=com"
"(objectClass=*)"
dn: uid=tmorris,ou=People,dc=example,dc=com
... snip ...
entryLevelRights: vadn
attributeLevelRights: givenName:rscwo, sn:rscwo, ou:rscwo, l:rscwo,
manager:rscwo, roomNumber:rscwo, mail:rscwo, facsimileTelephoneNumber:rscwo,
objectClass:rscwo, uid:rscwo, cn:rscwo, userPassword:rscwo
Only an administrator can retrieve the effective rights that a different user has on an entry. If T ed Morris
tried to determine Dave Miller's rights to Dave Miller's entry, then he would receive an insufficient access
error:
/usr/lib64/mozldap/ldapsearch -p 389 -h localhost -D
"uid=dmiller,ou=people,dc=example,dc=com" -w secret -b
"uid=tmorris,ou=people,dc=example,dc=com" -J
"1.3.6.1.4.1.42.2.27.9.5.2:true:dn:uid=tmorris,ou=people,dc=example,dc=com"
"(objectClass=*)"
ldap_search: Insufficient access
ldap_search: additional info: get-effective-rights: requester has no g permission
on the entry
However, a regular user can run a get effective rights search to see what rights another user has to his
personal entry. In Example 12.4, “Checking the Rights Someone Else Has to a Personal Entry”, T ed
Morris checks what rights Dave Miller has on T ed Morris's entry.
Example 12.4 . Checking the Rights Someone Else Has to a Personal Entry
/usr/lib64/mozldap/ldapsearch -p 389 -h localhost -D
"uid=tmorris,ou=people,dc=example,dc=com" -w secret -b
"uid=tmorris,ou=people,dc=example,dc=com" -J
"1.3.6.1.4.1.42.2.27.9.5.2:true:dn:uid=dmiller,ou=people,dc=example,dc=com"
"(objectClass=*)"
dn: uid=tmorris,ou=people,dc=example,dc=com
... snip ...
entryLevelRights: v
attributeLevelRights: givenName:rsc, sn:rsc, ou:rsc, l:rsc,manager:rsc,
roomNumber:rsc, mail:rsc, facsimileTelephoneNumber:rsc, objectClass:rsc,
uid:rsc, cn:rsc, userPassword:none
4 81
Red Hat D irectory Server 8.2 Administration Guide
In this case, Dave Miller has the right to view the DN of the entry and to read, search, and compare the
ou, givenName, l, and other attributes, and no rights to the userPassword attribute.
12.7.2.2. Examples of Get Effective Rights Searches for Non-Existent Attributes
By default, information is not given for attributes in an entry that do not have a value; for example, if the
userPassword value is removed, then a future effective rights search on the entry above would not
return any effective rights for userPassword, even though self-write and self-delete rights could be
allowed.
Using an asterisk (* ) with the get effective rights search returns every attribute available for the entry,
including attributes not set on the entry. For example:
Example 12.5. Returning Effective Rights for Non-Existent Attributes
/usr/lib64/mozldap/ldapsearch -D "cn=directory manager" -w secret12 -b
"uid=scarter,ou=people,dc=example,dc=com" -J
1.3.6.1.4.1.42.2.27.9.5.2:true:dn:uid=scarter,ou=people,dc=example,dc=com
"(objectclass=*)" "*"
dn: uid=scarter,ou=People,dc=example,dc=com
givenName: Sam
telephoneNumber: +1 408 555 4798
sn: Carter
ou: Accounting
ou: People
l: Sunnyvale
manager: uid=dmiller,ou=People,dc=example,dc=com
roomNumber: 4612
mail: [email protected]
facsimileTelephoneNumber: +1 408 555 9700
objectClass: top
objectClass: person
objectClass: organizationalPerson
objectClass: inetOrgPerson
uid: scarter
cn: Sam Carter
userPassword: {SSHA}Xd9Jt8g1UsHC8enNDrEmxj3iJPKQLItlDYdD9A==
entryLevelRights: vadn
attributeLevelRights: objectClass:rscwo, aci:rscwo, sn:rscwo, cn:rscwo,
description:rscwo, seeAlso:rscwo, telephoneNumber:rscwo, userPassword:rscwo,
destinationIndicator:rscwo, facsimileTelephoneNumber:rscwo,
internationaliSDNNumber:rscwo, l:rscwo, ou:rscwo,
physicalDeliveryOfficeName:rscwo, postOfficeBox:rscwo, postalAddress:rscwo,
postalCode:rscwo, preferredDeliveryMethod:rscwo, registeredAddress:rscwo,
st:rscwo, street:rscwo, teletexTerminalIdentifier:rscwo, telexNumber:rscwo,
title:rscwo, x121Address:rscwo, audio:rscwo, businessCategory:rscwo,
carLicense:rscwo, departmentNumber:rscwo, displayName:rscwo,
employeeType:rscwo, employeeNumber:rscwo, givenName:rscwo, homePhone:rscwo,
homePostalAddress:rscwo, initials:rscwo, jpegPhoto:rscwo, labeledUri:rscwo,
manager:rscwo, mobile:rscwo, pager:rscwo, photo:rscwo, preferredLanguage:rscwo,
mail:rscwo, o:rscwo, roomNumber:rscwo, secretary:rscwo,
uid:rscwo,x500UniqueIdentifier:rscwo, userCertificate:rscwo,
userSMIMECertificate:rscwo, userPKCS12:rscwo
All of the attributes available for the entry, such as secretary, are listed, even though that attribute is
non-existent.
4 82
Chapter 12. Managing Access Control
12.7.2.3. Examples of Get Effective Rights Searches for Specific Attributes or Object
Classes
T aking the attribute-related GER searches further, it is possible to search for the rights to a specific
attribute and set of attributes and to list all of the attributes available for one of the object classes set on
the entry.
One of the options listed in the formatting example in Section 12.7.2, “T he Format of a Get Effective
Rights Search” is attributeList. T o return the effective rights for only specific attributes, list the attributes,
separated by spaces, at the end of the search command. For example:
Example 12.6. Get Effective Rights Results for Specific Attributes
/usr/lib64/mozldap/ldapsearch -D "cn=directory manager" -w secret12 -b
"uid=scarter,ou=people,dc=example,dc=com" -J
1.3.6.1.4.1.42.2.27.9.5.2:true:dn:uid=scarter,ou=people,dc=example,dc=com
"(objectclass=*)" cn mail initials
dn: uid=scarter,ou=People,dc=example,dc=com
cn: Sam Carter
mail: [email protected]
entryLevelRights: vadn
attributeLevelRights: cn:rscwo, mail:rscwo, initials:rscwo
It is possible to specify a non-existent attribute in the attributeList, as with the initials attribute in
Example 12.6, “Get Effective Rights Results for Specific Attributes”, to see the rights which are available,
similar to using an asterisk to list all attributes.
T he Directory Manager can also list the rights for all of the attributes available to a specific object class.
T his option has the format attribute@objectClass. T his returns two entries; the first for the specified
GER subject and the second for a template entry for the object class. For example:
Example 12.7. Get Effective Rights Results for an Attribute within an Object Class
/usr/lib64/mozldap/ldapsearch -D "cn=directory manager" -w secret12 -b
"uid=scarter,ou=people,dc=example,dc=com" -J
1.3.6.1.4.1.42.2.27.9.5.2:true:dn:uid=scarter,ou=people,dc=example,dc=com
"(objectclass=*)" uidNumber@posixAccount
... snip ...
dn: cn=template_posixaccount_objectclass,uid=scarter,ou=people,dc=example,dc=com
uidnumber: (template_attribute)
entryLevelRights: v
attributeLevelRights: uidNumber:rsc
NOTE
Using the search format attribute@objectClass is only available if the requester (-D) is the
Directory Manager.
4 83
Red Hat D irectory Server 8.2 Administration Guide
If a regular user, rather than Directory Manager, tried to run the same command, the result would simply
be blank.
Example 12.8. Get Effective Rights Results with No ACL Set (Regular User)
$ /usr/lib64/mozldap/ldapsearch -D "uid=scarter,ou=people,dc=example,dc=com" w secret12 -b "dc=example,dc=com" -J
1.3.6.1.4.1.42.2.27.9.5.2:true:dn:uid=scarter,ou=people,dc=example,dc=com
"(objectclass=*)" "*@person"
$
Using an asterisk (* ) instead of a specific attribute returns all of the attributes (present and nonexistent) for the specified GER subject and the full list of attributes for the object class template. For
example:
Example 12.9. Get Effective Rights Results for All Attribute for an Object Class
/usr/lib64/mozldap/ldapsearch -D "cn=directory manager" -w secret12 -b
"uid=scarter,ou=people,dc=example,dc=com" -J
1.3.6.1.4.1.42.2.27.9.5.2:true:dn:uid=scarter,ou=people,dc=example,dc=com
"(objectclass=*)" *@posixaccount
... snip ...
dn: cn=template_posixaccount_objectclass,uid=scarter,ou=people,dc=example,dc=com
objectClass: posixaccount
objectClass: top
homeDirectory: (template_attribute)
gidNumber: (template_attribute)
uidNumber: (template_attribute)
uid: (template_attribute)
cn: (template_attribute)
entryLevelRights: v
attributeLevelRights: cn:rsc, uid:rsc, uidNumber:rsc, gidNumber:rsc,
homeDirectory:rsc, objectClass:rsc, userPassword:none, loginShell:rsc,
gecos:rsc, description:rsc, aci:rsc
12.7.2.4 . Examples of Get Effective Rights Searches for Operational Attributes
Operational attributes are not returned in regular ldapsearches, including get effective rights
searches. T o return the information for the operational attributes, use the plus sign (+). T his returns
only the operational attributes that can be used in the entry. For example:
4 84
Chapter 12. Managing Access Control
Example 12.10. Get Effective Rights Results for Operational Attributes
/usr/lib64/mozldap/ldapsearch -D "cn=directory manager" -w secret12 -b
"uid=scarter,ou=people,dc=example,dc=com" -J
1.3.6.1.4.1.42.2.27.9.5.2:true:dn:uid=scarter,ou=people,dc=example,dc=com
"(objectclass=*)" "+"
dn: uid=scarter,ou=People,dc=example,dc=com
entryLevelRights: vadn
attributeLevelRights: nsICQStatusText:rscwo, passwordGraceUserTime:rscwo,
pwdGraceUserTime:rscwo, nsYIMStatusText:rscwo, modifyTimestamp:rscwo,
passwordExpWarned:rscwo, pwdExpirationWarned:rscwo, entrydn:rscwo, aci:rscwo,
nsSizeLimit:rscwo, nsAccountLock:rscwo, passwordExpirationTime:rscwo,
entryid:rscwo, nsSchemaCSN:rscwo, nsRole:rscwo, retryCountResetTime:rscwo,
ldapSchemas:rscwo, nsAIMStatusText:rscwo, copiedFrom:rscwo,
nsICQStatusGraphic:rscwo, nsUniqueId:rscwo, creatorsName:rscwo,
passwordRetryCount:rscwo, dncomp:rscwo, nsTimeLimit:rscwo,
passwordHistory:rscwo, pwdHistory:rscwo, nscpEntryDN:rscwo,
subschemaSubentry:rscwo, nsYIMStatusGraphic:rscwo, hasSubordinates:rscwo,
pwdpolicysubentry:rscwo, nsAIMStatusGraphic:rscwo, nsRoleDN:rscwo,
createTimestamp:rscwo, accountUnlockTime:rscwo, copyingFrom:rscwo,
nsLookThroughLimit:rscwo, nsds5ReplConflict:rscwo, modifiersName:rscwo,
parentid:rscwo, passwordAllowChangeTime:rscwo, nsBackendSuffix:rscwo,
nsIdleTimeout:rscwo, ldapSyntaxes:rscwo, numSubordinates:rscwo
12.7.2.5. Examples of Get Effective Rights Results and Access Control Rules
Get effective rights are returned according to to whatever ACLs are in effect for the GER subject entry.
For example, this ACL is set and, for the purposes of this example, it is the only ACL set:
dn: dc=example,dc=com
objectClass: top
objectClass: domain
dc: example
aci: (target=ldap:///ou=Accounting,dc=example,dc=com)(targetattr="*")(version
3.0; acl "test acl"; allow (read,search,compare) (userdn = "ldap:///anyone") ;)
dn: ou=Accounting,dc=example,dc=com
objectClass: top
objectClass: organizationalUnit
ou: Accounting
Because the ACL does not include the dc=exam ple,dc=com subtree, the get effective rights search
shows that the user does not have any rights to the dc=exam ple,dc=com entry:
4 85
Red Hat D irectory Server 8.2 Administration Guide
Example 12.11. Get Effective Rights Results with No ACL Set (Directory Manager)
/usr/lib64/mozldap/ldapsearch -D "cn=directory manager" -w secret12 -b
"dc=example,dc=com" -J
1.3.6.1.4.1.42.2.27.9.5.2:true:dn:uid=scarter,ou=people,dc=example,dc=com
"(objectclass=*)" "*@person"
dn: cn=template_person_objectclass,uid=scarter,ou=people,dc=example,dc=com
objectClass: person
objectClass: top
cn: (template_attribute)
sn: (template_attribute)
description: (template_attribute)
seeAlso: (template_attribute)
telephoneNumber: (template_attribute)
userPassword: (template_attribute)
entryLevelRights: none
attributeLevelRights: sn:none, cn:none, objectClass:none, description:none,
seeAlso:none, telephoneNumber:none, userPassword:none, aci:none
12.7.3. Using Get Effective Rights from the Console
T o view effective rights from the Console:
1. Open the Directory tab, and right-click the entry of which to check the rights.
2. Select Advanced Properties from the drop-down menu.
3. Check the Show effective rights checkbox.
4. Beside each attribute, the attribute-level get effective rights are displayed. T he entry-level rights
are shown beneath the entry's DN.
4 86
Chapter 12. Managing Access Control
T he attribute-level effective rights (r, s, c, w, o) appear next to the attributes. T he entry-level
rights (v, a, d, n) appear under the full DN for the entry in the lower left-hand corner of the
Property Editor.
If you check the Show all allowed attributes checkbox, then the effective rights for those
attributes appear next to the additional attributes, even though they do not have values.
12.7.4. Get Effective Rights Return Codes
If the criticality is set to false for a get effective rights search and an error occurs, the regular entry
information is returned, but, in place of rights for entryLevelRights and attributeLevelRights,
an error code is returned. T his code can give information on the configuration of the entry that was
queried. T able 12.8, “Returned Result Codes” summarizes the error codes and the potential
configuration information they can relay.
4 87
Red Hat D irectory Server 8.2 Administration Guide
T able 12.8. Returned Result Codes
Code
Description
0
Successfully completed.
1
Operation error.
12
T he critical extension is unavailable. If the
criticality expression is set to true and effective
rights do not exist on the entry being queried,
then this error is returned.
16
No such attribute. If an attribute is specifically
queried for access rights but that attribute does
not exist in the schema, this error is returned.
17
Undefined attribute type.
21
Invalid attribute syntax.
50
Insufficient rights.
52
Unavailable.
53
Unwilling to perform.
80
Other.
12.8. Logging Access Control Information
T o obtain information on access control in the error logs, you must set the appropriate log level. T o set
the error log level from the Console:
1. In the Console, click the Directory tab, right-click the config node, and choose Properties
from the pop-up menu.
T his displays the Property Editor for the cn=config entry.
2. Scroll down the list of attribute value pairs to locate the nsslapd-errorlog-level attribute.
3. Add 128 to the value already displayed in the nsslapd-errorlog-level value field.
For example, if the value already displayed is 8192 (replication debugging), change the value to
8320. For complete information on error log levels, see the Directory Server Configuration and
Command-Line Tool Reference.
4 88
Chapter 12. Managing Access Control
4. Click OK to dismiss the Property Editor.
12.9. Access Control Usage Examples
T he examples provided in this section illustrate how an imaginary ISP company, Example Corp., would
implement its access control policy. All the examples explain how to perform a given task from the
Console and using an LDIF file.
Example Corp.'s business is to offer a web hosting service and Internet access. Part of Example Corp.'s
web hosting service is to host the directories of client companies. Example Corp. actually hosts and
partially manages the directories of two medium-sized companies, HostedCom pany1 and
HostedCom pany2. It also provides Internet access to a number of individual subscribers.
T hese are the access control rules that Example Corp. wants to put in place:
Grant anonymous access for read, search, and compare to the entire exam ple tree for Example
Corp. employees (Section 12.9.1, “Granting Anonymous Access”).
Grant write access to Example Corp. employees for personal information, such as homePhone and
homePostalAddress (Section 12.9.2, “Granting Write Access to Personal Entries”).
Grant Example Corp. employees the right to add any role to their entry, except certain critical roles
(Section 12.9.3, “Restricting Access to Key Roles”).
Grant the exam ple.com Hum an Resources group all rights on the entries in the People branch
(Section 12.9.4, “Granting a Group Full Access to a Suffix”).
Grant all Example Corp. employees the right to create group entries under the Social Com m ittee
branch of the directory and to delete group entries that they own (Section 12.9.5, “Granting Rights to
Add and Delete Group Entries”).
Grant all Example Corp. employees the right to add themselves to group entries under the Social
Com m ittee branch of the directory (Section 12.9.9, “Allowing Users to Add or Remove T hemselves
from a Group”).
Grant access to the directory administrator (role) of HostedCom pany1 and HostedCom pany2 on
their respective branches of the directory tree, with certain conditions such as SSL authentication,
time and date restrictions, and specified location (Section 12.9.6, “Granting Conditional Access to a
Group or Role”).
Deny individual subscribers access to the billing information in their own entries (Section 12.9.7,
“Denying Access”).
Grant anonymous access to the world to the individual subscribers subtree, except for subscribers
4 89
Red Hat D irectory Server 8.2 Administration Guide
who have specifically requested to be unlisted. (T his part of the directory could be a consumer
server outside of the firewall and be updated once a day.) See Section 12.9.1, “Granting Anonymous
Access” and Section 12.9.8, “Setting a T arget Using Filtering”.
12.9.1. Granting Anonymous Access
Most directories are run such that you can anonymously access at least one suffix for read, search, or
compare. For example, you might want to set these permissions if you are running a corporate personnel
directory that you want employees to be able to search, such as a phonebook. T his is the case at
Example Corp. internally and is illustrated in Section 12.9.1.1, “ACI "Anonymous example.com"”.
As an ISP, Example Corp. also wants to advertise the contact information of all of its subscribers by
creating a public phonebook accessible to the world. T his is illustrated in Section 12.9.1.2, “ACI
"Anonymous World"”.
12.9.1.1. ACI "Anonymous example.com"
In LDIF, to grant read, search, and compare permissions to the entire Example Corp. tree to Example
Corp. employees, write the following statement:
aci: (targetattr !="userPassword")(version 3.0; acl "Anonymous
Example"; allow (read, search, compare) userdn= "ldap:///anyone"
and dns="*.example.com";)
T his example assumes that the aci attribute is added to the dc=exam ple,dc=com entry. T he
userPassword attribute is excluded from the scope of the ACI.
From the Console:
1. In the Directory tab, right-click the exam ple node in the left navigation tree, and choose Set
Access Perm issions from the pop-up menu to display the Access Control Manager.
2. Click New to display the Access Control Editor.
3. In the Users/Groups tab in the ACI nam e field, type Anonym ous exam ple.com . Check that
All Users opens in the list of users granted access permission.
4. In the Rights tab, select the checkboxes for read, com pare, and search rights. Make sure the
other checkboxes are clear.
5. In the T argets tab, click T his Entry to display the dc=exam ple,dc=com suffix in the
T arget directory entry field. In the attribute table, locate the userPassword attribute, and
4 90
Chapter 12. Managing Access Control
clear the corresponding checkbox.
All other checkboxes should be selected. T his task is easier if you click the Nam e header to
organize the list of attributes alphabetically.
6. In the Hosts tab, click Add, and in the DNS host filter field, type * .exam ple.com . Click OK
to dismiss the dialog box.
7. Click OK in the Access Control Editor window.
12.9.1.2. ACI "Anonymous World"
In LDIF, to grant read and search access of the individual subscribers subtree to the world, while
denying access to information on unlisted subscribers, write the following statement:
aci: (targetfilter= "(!(unlistedSubscriber=yes))")
(targetattr="homePostalAddress || homePhone || mail") (version
3.0; acl "Anonymous World"; allow (read, search) userdn="ldap:///anyone";)
T his example assumes that the ACI is added to the ou=subscribers,dc=exam ple,dc=com entry. It
also assumes that every subscriber entry has an unlistedSubscriber attribute which is set to yes
or no. T he target definition filters out the unlisted subscribers based on the value of this attribute. For
details on the filter definition, see Section 12.9.8, “Setting a T arget Using Filtering”.
From the Console:
1. In the Directory tab, right-click the Subscribers entry under the exam ple node in the left
navigation tree, and choose Set Access Perm issions from the pop-up menu to display the
Access Control Manager.
2. Click New to display the Access Control Editor.
3. In the Users/Groups tab, in the ACI nam e field, type Anonym ous World. Check that All
Users opens in the list of users granted access permission.
4 91
Red Hat D irectory Server 8.2 Administration Guide
4. In the Rights tab, select the checkboxes for read and search rights. Make sure the other
checkboxes are clear.
5. In the T argets tab, click T his Entry to display the
ou=subscribers,dc=exam ple,dc=com suffix in the T arget directory entry field.
6. In the Filter for subentries field, enter a filter which excludes unlisted subscribers
((!(unlistedSubscriber=yes))).
7. In the attribute table, select the checkboxes for the homePhone, homePostalAddress, and mail
attributes.
All other checkboxes should be clear; if it is easier, click the Check None button to clear the
checkboxes for all attributes in the table, then click the Nam e header to organize them
alphabetically, and select the appropriate ones.
8. Click OK.
12.9.2. Granting Write Access to Personal Entries
Many directory administrators want to allow internal users to change some but not all of the attributes in
their own entry. T he directory administrators at Example Corp. want to allow users to change their own
password, home telephone number, and home address, but nothing else. T his is illustrated in
Section 12.9.2.1, “ACI "Write example.com"”.
It is also Example Corp.'s policy to let their subscribers update their own personal information in the
exam ple tree, provided that they establish an SSL connection to the directory. T his is illustrated in
Section 12.9.2.2, “ACI "Write Subscribers"”.
12.9.2.1. ACI "Write example.com"
4 92
Chapter 12. Managing Access Control
NOTE
By setting this permission, users will also have the right to delete attribute values.
Granting Example Corp. employees the right to update their password, home telephone number, and
home address has the following statement in LDIF:
aci: (targetattr="userPassword || homePhone ||
homePostalAddress") (version 3.0; acl "Write example.com"; allow
(write) userdn= "ldap:///self" and dns="*.example.com";)
T his example assumes that the ACI is added to the ou=people,dc=exam ple,dc=com entry.
From the Console:
1. In the Directory tab, right-click the people entry under the exam ple node in the left
navigation tree, and choose Set Access Perm issions from the pop-up menu to display the
Access Control Manager.
2. Click New to display the Access Control Editor.
3. In the Users/Groups tab, in the ACI nam e field, type Write exam ple.com . In the list of users
granted access permission:
a. Remove All Users from the targeted user list.
b. Click Add.
c. Set the Search area to Special Rights, and select Self from the search results list.
d. Click the Add button to list Self in the list of users who are granted access permission.
4 93
Red Hat D irectory Server 8.2 Administration Guide
e. Click OK to dismiss the Add Users and Groups dialog box.
4. In the Rights tab, select the checkbox for write right. Make sure the other checkboxes are
clear.
5. In the T argets tab, click T his Entry to display the ou=people,dc=exam ple,dc=com suffix
in the T arget directory entry field. In the attribute table, select the checkboxes for the
homePhone, homePostalAddress, and userPassword attributes.
4 94
Chapter 12. Managing Access Control
All other checkboxes should be clear; if it is easier, click the Check None button to clear the
checkboxes for all attributes in the table, then click the Nam e header to organize them
alphabetically, and select the appropriate ones.
6. In the Hosts tab, click Add to display the Add Host Filter dialog box. In the DNS host
filter field, type * .exam ple.com . Click OK to dismiss the dialog box.
7. Click OK in the Access Control Editor window.
12.9.2.2. ACI "Write Subscribers"
4 95
Red Hat D irectory Server 8.2 Administration Guide
NOTE
By setting this permission, you are also granting users the right to delete attribute values.
In LDIF, to grant Example Corp. subscribers the right to update their password and home telephone
number, write the following statement:
aci: (targetattr = "homePhone || homePostalAddress || mail")
(target = "ldap:///ou=Subscribers,dc=example,dc=com")
(targetfilter = (!(unlistedSubscriber=yes)) )
(version 3.0;
acl "Write Subscribers";
allow (write)
(userdn = "ldap:///self") and authmethod="ssl";
;)
T his example assumes that the aci is added to the ou=subscribers,dc=exam ple,dc=com entry.
Example Corp. subscribers do not have write access to their home address because they might delete
the attribute, and Example Corp. needs that information for billing. T herefore, the home address is
business-critical information.
From the Console:
1. In the Directory tab, right-click the subscribers entry under the exam ple node in the left
navigation tree, and choose Set Access Perm issions from the pop-up menu to display the
Access Control Manager.
2. Click New to display the Access Control Editor.
3. In the Users/Groups tab, in the ACI nam e field, type Write Subscribers. In the list of users
granted access permission:
a. Remove All Users from the targeted user list.
b. Click Add.
c. Set the Search area to Special Rights, and select Self from the search results list.
d. Click the Add button to list Self in the list of users who are granted access permission.
e. Click OK to dismiss the Add Users and Groups dialog box.
4. In the Rights tab, select the checkbox for write. Make sure the other checkboxes are clear.
5. In the T argets tab, click T his Entry to display the
ou=subscribers,dc=exam ple,dc=com suffix in the T arget directory entry field.
a. In the Filter for subentries field, type a filter so that only listed subscribers are
included:
(!(unlistedSubscriber=yes))
b. In the attribute table, select the checkboxes for the homePhone, homePostalAddress, and
mail attributes.
4 96
Chapter 12. Managing Access Control
All other checkboxes should be clear; if necessary, click the Check None button to clear
the checkboxes for all attributes in the table, then click the Nam e header to organize them
alphabetically, and select the appropriate ones.
c. Optionally, to require users to authenticate using SSL, switch to manual editing by clicking
the Edit Manually button, and add authm ethod=ssl to the LDIF statement:
(targetattr = "homePhone || homePostalAddress || mail")
(target = "ldap:///ou=Subscribers,dc=example,dc=com")
(targetfilter = (!(unlistedSubscriber=yes)) )
(version 3.0;
acl "Write Subscribers";
allow (write)
(userdn = "ldap:///self") and authmethod="ssl";
;)
6. Click OK.
12.9.3. Restricting Access to Key Roles
You can use role definitions in the directory to identify functions that are critical to your business, the
administration of your network and directory, or another purpose.
For example, you might create a superAdm in role by identifying a subset of your system administrators
that are available at a particular time of day and day of the week at corporate sites worldwide, or you
might want to create a First Aid role that includes all members of staff on a particular site that have
done first aid training. For information on creating role definitions, see Section 11.2, “Using Roles”.
When a role gives any sort of privileged user rights over critical corporate or business functions,
consider restricting access to that role. For example, at Example Corp., employees can add any role to
their own entry except the superAdm in role. T his is illustrated in Section 12.9.3.1, “ACI "Roles"”.
4 97
Red Hat D irectory Server 8.2 Administration Guide
12.9.3.1. ACI "Roles"
In LDIF, to grant Example Corp. employees the right to add any role to their own entry except the
superAdm in role, write the following statement:
aci: (targetattr = "nsroledn")
(targattrfilters="add=nsroledn:(nsroledn !=
"cn=superAdmin,dc=example,dc=com")") (version 3.0; acl "Roles";
allow (write) userdn= "ldap:///self" and dns="*.example.com";)
T his example assumes that the ACI is added to the ou=people,dc=exam ple,dc=com entry.
From the Console:
1. In the Directory tab, right-click the people entry under the exam ple node in the left
navigation tree, and choose Set Access Perm issions from the pop-up menu to display the
Access Control Manager.
2. Click New to display the Access Control Editor.
3. In the Users/Groups tab, in the ACI nam e field, type Roles. In the list of users granted access
permission:
a. Remove All Users from the targeted user list.
b. Click Add.
c. Set the Search area in the Add Users and Groups dialog box to Special Rights,
and select Self from the search results list.
d. Click the Add button to list Self in the list of users who are granted access permission.
e. Click OK to dismiss the Add Users and Groups dialog box.
4. In the Rights tab, select the checkbox for write. Make sure the other checkboxes are clear.
5. In the T argets tab, click T his Entry to use the ou=people,dc=exam ple,dc=com suffix in
the T arget directory entry field.
6. In the Hosts tab, click Add to display the Add Host Filter dialog box. In the DNS host
filter field, type * .exam ple.com . Click OK to dismiss the dialog box.
7. T o create the value-based filter for roles, switch to manual editing by clicking the Edit
Manually button. Add the following to the beginning of the LDIF statement:
(targattrfilters="add=nsroledn:(nsroledn !=
"cn=superAdmin,dc=example,dc=com")")
T he LDIF statement should read as follows:
(targattrfilters="add=nsroledn:(nsroledn != "cn=superAdmin,
dc=example,dc=com")") (targetattr = "*") (target = "ldap:///
ou=people,dc=example,dc=com") (version 3.0; acl "Roles";
allow (write) (userdn = "ldap:///self") and (dns="*.example.com");)
8. Click OK.
12.9.4. Granting a Group Full Access to a Suffix
Most directories have a group that is used to identify certain corporate functions. T hese groups can be
given full access to all or part of the directory. By applying the access rights to the group, you can avoid
setting the access rights for each member individually. Instead, you grant users these access rights
simply by adding them to the group.
4 98
Chapter 12. Managing Access Control
For example, when the Directory Server is set up with a typical process, an administrators group with full
access to the directory is created by default.
At Example Corp., the Hum an Resources group is allowed full access to the ou=people branch of the
directory so that they can update the employee database. T his is illustrated in Section 12.9.4.1, “ACI
"HR"”.
12.9.4 .1. ACI "HR"
In LDIF, to grant the HR group all rights on the employee branch of the directory, use the following
statement:
aci: (version 3.0; acl "HR"; allow (all) userdn=
"ldap:///cn=HRgroup,ou=people,dc=example,dc=com";)
T his example assumes that the ACI is added to the ou=people,dc=exam ple,dc=com entry.
From the Console:
1. In the Directory tab, right-click the people entry under the exam ple node in the left
navigation tree, and choose Set Access Perm issions from the pop-up menu to display the
Access Control Manager.
2. Click New to display the Access Control Editor.
3. In the Users/Groups tab, in the ACI nam e field, type HR. In the list of users granted access
permission:
a. Remove All Users from the targeted user list.
b. Click Add.
c. Set the Search area to Users and Groups, and type HRgroup in the Search for
field.
4 99
Red Hat D irectory Server 8.2 Administration Guide
T his example assumes that you have created an HR group or role. For more information on
groups and roles, see Chapter 11, Organizing Entries with Groups, Roles, and Views.
d. Click the Add button to list the HR group in the list of users who are granted access
permission.
e. Click OK to dismiss the Add Users and Groups dialog box.
4. In the Rights tab, click the Check All button.
All checkboxes are selected, except for proxy rights.
500
Chapter 12. Managing Access Control
5. Click OK.
12.9.5. Granting Rights to Add and Delete Group Entries
Some organizations want to allow employees to create entries in the tree if it can increase their
efficiency or if it can contribute to the corporate dynamics.
At Example Corp., there is an active social committee that is organized into various clubs, such as
tennis, swimming, and skiing. Any Example Corp. employee can create a group entry representing a new
club. T his is illustrated in Section 12.9.5.1, “ACI "Create Group"”. Any Example Corp. employee can
become a member of one of these groups. T his is illustrated in Section 12.9.9.1, “ACI "Group Members"”
under Section 12.9.9, “Allowing Users to Add or Remove T hemselves from a Group”. Only the group
owner can modify or delete a group entry. T his is illustrated in Section 12.9.5.2, “ACI "Delete Group"”.
12.9.5.1. ACI "Create Group"
In LDIF, to grant Example Corp. employees the right to create a group entry under the ou=Social
Com m ittee branch, write the following statement:
aci: (target="ldap:///ou=social committee,dc=example,dc=com)
(targattrfilters="add=objectClass:(objectClass=groupOfNames)")
(version 3.0; acl "Create Group"; allow (add)
(userdn= "ldap:///uid=*,ou=people,dc=example,dc=com")
and dns="*.example.com";)
NOTE
T his ACI does not grant write permission, which means that the entry creator cannot modify the
entry.
T his example assumes that the ACI is added to the ou=social com m ittee,dc=exam ple,dc=com
entry.
From the Console:
1. In the Directory tab, right-click the Social Com m ittee entry under the exam ple node in the
left navigation tree, and choose Set Access Perm issions from the pop-up menu to display
the Access Control Manager.
2. Click New to display the Access Control Editor.
3. In the Users/Groups tab, in the ACI nam e field, type Create Group. In the list of users
granted access permission:
a. Remove All Users from the targeted user list.
b. Click Add.
c. Set the Search area to Special Rights, and select All Authenticated Users
from the search results list.
d. Click the Add button to list All Authenticated Users in the list of users who are
granted access permission.
501
Red Hat D irectory Server 8.2 Administration Guide
e. Click OK to dismiss the Add Users and Groups dialog box.
4. In the Rights tab, select the checkboxes for add, search, and read. Make sure the other
checkboxes are clear.
5. In the T argets tab, click T his Entry to display the ou=social
502
Chapter 12. Managing Access Control
com m ittee,dc=exam ple,dc=com suffix in the T arget directory entry field.
6. In the Hosts tab, click Add to display the Add Host Filter dialog box. In the DNS host
filter field, type * .exam ple.com . Click OK to dismiss the dialog box.
7. T o create the value-based filter that allows employees to add only group entries to this subtree,
click the Edit Manually button. Add the following to the beginning of the LDIF statement:
(targattrfilters="add=objectClass:(objectClass=groupOfNames)")
T he LDIF statement should read as follows:
(targattrfilters="add=objectClass:(objectClass=groupOfNames)")
(targetattr = "*") (target="ldap:///ou=social
committee,dc=example,dc=com)
(version 3.0; acl "Create Group"; allow (read,search,add)
(userdn= "ldap:///all") and (dns="*.example.com"); )
8. Click OK.
12.9.5.2. ACI "Delete Group"
In LDIF, to grant Example Corp. employees the right to modify or delete a group entry which they own
under the ou=Social Com m ittee branch, write the following statement:
aci: (target="ou=social committee,dc=example,dc=com)
(targattrfilters="del=objectClass:(objectClass=groupOfNames)")
(version 3.0; acl "Delete Group"; allow (delete) userattr=
"owner#GROUPDN";)
T his example assumes that the aci is added to the ou=social com m ittee,dc=exam ple,dc=com
entry.
NOTE
Using the Console is not an effective way of creating this ACI because it requires manually editing
the ACI to create the target filter and to check group ownership.
12.9.6. Granting Conditional Access to a Group or Role
In many cases, when you grant a group or role privileged access to the directory, you want to ensure
that those privileges are protected from intruders trying to impersonate your privileged users. T herefore,
in many cases, access control rules that grant critical access to a group or role are often associated
with a number of conditions.
Example Corp. has created a directory administrator role for each of its hosted companies,
HostedCom pany1 and HostedCom pany2. It wants these companies to be able to manage their own
data and implement their own access control rules while securing it against intruders. For this reason,
HostedCom pany1 and HostedCom pany2 have full rights on their respective branches of the directory
tree, provided the following conditions are fulfilled:
Connection authenticated using SSL
Access requested between 8 a.m. and 6 p.m., Monday through T hursday
Access requested from a specified IP address for each company
503
Red Hat D irectory Server 8.2 Administration Guide
T hese conditions are illustrated in a single ACI for each company, HostedCom pany1 and
HostedCom pany2. Because the content of these ACIs is the same, the examples below illustrate the
HostedCom pany1 ACI only.
12.9.6.1. ACI "HostedCompany1"
In LDIF, to grant HostedCom pany1 full access to their own branch of the directory under the requisite
conditions, write the following statement:
aci:(target="ou=HostedCompany1,ou=corporate-clients,dc=example,dc=com")
(targetattr= "*") (version 3.0; acl "HostedCompany1";allow (all)
(roledn="ldap:///cn=DirectoryAdmin,ou=HostedCompany1,
ou=corporate-clients,dc=example,dc=com") and
(authmethod="ssl") and (dayofweek="Mon,Tues,Wed,Thu") and (timeofday >= "0800"
and
timeofday <= "1800") and (ip="255.255.123.234"); )
T his example assumes that the ACI is added to the ou=HostedCom pany1,ou=corporateclients,dc=exam ple,dc=com entry.
From the Console:
1. In the Directory tab, right-click the HostedCom pany1 entry under the exam ple node in the
left navigation tree, and choose Set Access Perm issions from the pop-up menu to display
the Access Control Manager.
2. Click New to display the Access Control Editor.
3. In the Users/Groups tab, type HostedCom pany1 in the ACI nam e field. In the list of users
granted access permission:
a. Remove All Users from the targeted user list.
b. Click Add.
c. Set the Search area to Users and Groups, and type DirectoryAdm in in the Search
For field.
(T his assumes that there is an administrator's role with a cn of DirectoryAdm in.)
d. Click the Add button to list the administrator's role in the list of users who are granted
access permission.
e. Click OK to dismiss the Add Users and Groups dialog box.
4. In the Rights tab, click the Check All button.
5. In the T argets tab, click T his Entry to display the ou=HostedCom pany1,ou=corporateclients,dc=exam ple,dc=com suffix in the T arget directory entry field.
6. In the Hosts tab, click Add to display the Add Host Filter dialog box. In the IP address
host filter field, type 255.255.123.234 . Click OK.
504
Chapter 12. Managing Access Control
T he IP address must be a valid IP address for the host machine that the HostedCom pany1
administrators use to connect to the exam ple directory.
7. In the T im es tab, select the block time corresponding to Monday through T hursday and 8 a.m. to
6 p.m.
A message appears below the table that specifies the selected time block.
8. T o enforce SSL authentication from HostedCom pany1 administrators, switch to manual editing
by clicking the Edit Manually button. Add the following to the end of the LDIF statement:
and (authmethod="ssl")
T he LDIF statement should be similar to the following:
(targetattr = "*") (target="ou=HostedCompany1,ou=corporateclients,dc=example,dc=com")
(version 3.0; acl "HostedCompany1"; allow (all) (roledn=
"ldap:///cn=DirectoryAdmin,ou=HostedCompany1,ou=corporateclients,dc=example,dc=com") and
(dayofweek="Mon,Tues,Wed,Thu") and (timeofday >= "0800" and timeofday <=
"1800") and
(ip="255.255.123.234") and (authmethod="ssl"); )
9. Click OK.
505
Red Hat D irectory Server 8.2 Administration Guide
Click OK.
12.9.7. Denying Access
If your directory holds business-critical information, it may be necessary to specifically deny access to it.
For example, Example Corp. wants all subscribers to be able to read billing information such as
connection time or account balance under their own entries but explicitly wants to deny write access to
that information. T his is illustrated in Section 12.9.7.1, “ACI "Billing Info Read"” and Section 12.9.7.2, “ACI
"Billing Info Deny"”, respectively.
12.9.7.1. ACI "Billing Info Read"
In LDIF, to grant subscribers permission to read billing information in their own entry, write the following
statement:
aci: (targetattr="connectionTime || accountBalance") (version
3.0; acl "Billing Info Read"; allow (search,read) userdn=
"ldap:///self";)
T his example assumes that the relevant attributes have been created in the schema and that the ACI is
added to the ou=subscribers,dc=exam ple,dc=com entry.
From the Console:
1. In the Directory tab, right-click the Subscribers entry under the exam ple node in the left
navigation tree, and choose Set Access Perm issions from the pop-up menu to display the
Access Control Manager.
2. Click New to display the Access Control Editor.
3. In the Users/Groups tab, in the ACI nam e field, type Billing Info Read. In the list of users
granted access permission:
a. Remove All Users from the targeted user list.
b. Click Add.
c. Set the Search area in the Add Users and Groups dialog box to Special Rights,
and select Self from the search results list.
d. Click the Add button to list Self in the list of users who are granted access permission.
e. Click OK to dismiss the Add Users and Groups dialog box.
4. In the Rights tab, select the checkboxes for search and read rights. Make sure the other
checkboxes are clear.
506
Chapter 12. Managing Access Control
5. In the T argets tab, click T his Entry to display the
ou=subscribers,dc=exam ple,dc=com suffix in the T arget directory entry field. In
the attribute table, select the checkboxes for the connectionTime and accountBalance
attributes. (T hese are custom schema that Example Corp. uses for ISP account management.)
All other checkboxes should be clear; if it is easier, click the Check None button to clear the
checkboxes for all attributes in the table, then click the Nam e header to organize them
alphabetically, and select the appropriate ones.
T his example assumes that you have added the connectionTime and accountBalance
attributes to the schema.
6. Click OK.
12.9.7.2. ACI "Billing Info Deny"
In LDIF, to deny subscribers permission to modify billing information in their own entry, write the following
statement:
aci: (targetattr="connectionTime || accountBalance") (version
3.0; acl "Billing Info Deny"; deny (write) userdn="ldap:///self";)
T his example assumes that the relevant attributes have been created in the schema and that the ACI is
added to the ou=subscribers,dc=exam ple,dc=com entry.
From the Console:
1. In the Directory tab, right-click the Subscribers entry under the exam ple node in the left
navigation tree, and choose Set Access Perm issions from the pop-up menu to display the
Access Control Manager.
2. Click New to display the Access Control Editor.
3. In the Users/Groups tab, in the ACI nam e field, type Billing Info Deny. In the list of users
granted access permission:
a. Remove All Users from the targeted user list.
b. Click Add.
c. Set the Search area in the Add Users and Groups dialog box to Special Rights,
and select Self from the search results list.
d. Click the Add button to list Self in the list of users who are granted access permission.
e. Click OK to dismiss the Add Users and Groups dialog box.
4. In the Rights tab, select the checkbox for write. Make sure the other checkboxes are clear.
5. Click the Edit Manually button, and, in the LDIF statement that opens, change the word allow
to deny.
507
Red Hat D irectory Server 8.2 Administration Guide
6. In the T argets tab, click T his Entry to display the
ou=subscribers,dc=exam ple,dc=com suffix in the T arget directory entry field. In
the attribute table, select the checkboxes for the connectionTime and accountBalance
attributes.
All other checkboxes should be clear; if it is easier, click the Check None button to clear the
checkboxes for all attributes in the table, then click the Nam e header to organize them
alphabetically, and select the appropriate ones.
T his example assumes that the connectionTime and accountBalance attributes were added to
the schema.
7. Click OK.
12.9.8. Setting a Target Using Filtering
T o set access controls that allow access to a number of entries that are spread across the directory,
consider using a filter to set the target.
NOTE
Because search filters do not directly name the object for which you are managing access, it is
easy to allow or deny access to the wrong objects unintentionally, especially as your directory
becomes more complex. Additionally, filters can make it difficult to troubleshoot access control
problems within your directory.
For example, the following ACI grants user bjensen write access to the department number, home
phone number, home postal address, and manager attributes for all members of the accounting
organization.
aci: (targetattr="departmentNumber || homePhone || homePostalAddress || manager")
(targetfilter="(uid=bjensen)") (version 3.0; acl "Filtered ACL"; allow (write)
userdn ="ldap:///cn=*,ou=accounting,dc=example,dc=com";)
Before you can set these permissions, you must create the accounting branch point
ou=accounting,dc=exam ple,dc=com ). You can create organizational unit branch points in the
Directory tab on the Directory Server Console.
12.9.9. Allowing Users to Add or Remove Themselves from a Group
Many directories set ACIs that allow users to add or remove themselves from groups. T his is useful, for
example, for allowing users to add and remove themselves from mailing lists.
At Example Corp., employees can add themselves to any group entry under the ou=social
508
Chapter 12. Managing Access Control
com m ittee subtree. T his is illustrated in Section 12.9.9.1, “ACI "Group Members"”.
12.9.9.1. ACI "Group Members"
In LDIF, to grant Example Corp. employees the right to add or delete themselves from a group, write the
following statement:
aci: (targettattr="member")(version 3.0; acl "Group Members"; allow (selfwrite)
(userdn= "ldap:///uid=*,ou=people,dc=example,dc=com") ;)
T his example assumes that the ACI is added to the ou=social com m ittee,dc=exam ple,dc=com
entry.
From the Console:
1. In the Directory tab, right-click the people entry under the exam ple node in the left
navigation tree, and choose Set Access Perm issions from the pop-up menu to display the
Access Control Manager.
2. Click New to display the Access Control Editor.
3. In the Users/Groups tab, in the ACI nam e field, type Group Mem bers. In the list of users
granted access permission:
a. Remove All Users from the targeted user list.
b. Click Add.
c. Set the Search area in the Add Users and Groups dialog box to Special Rights,
and select All Authenticated Users from the search results list.
d. Click the Add button to list All Authenticated Users in the list of users who are
granted access permission.
e. Click OK to dismiss the Add Users and Groups dialog box.
4. In the Rights tab, select the checkbox for selfwrite. Make sure the other checkboxes are
clear.
5. In the T argets tab, type dc=exam ple,dc=com suffix in the T arget directory entry field.
In the attribute table, select the checkbox for the member attribute.
509
Red Hat D irectory Server 8.2 Administration Guide
All other checkboxes should be clear; if it is easier, click the Check None button to clear the
checkboxes for all attributes in the table, then click the Nam e header to organize them
alphabetically, and select the appropriate ones.
6. Click OK.
12.9.10. Setting an ACI to Require a Certain Security Strength Factor for Some
Operations
As mentioned in Section 12.4.8, “Requiring a Certain Level of Security in Connections”, the ssf keyword
is used to require a secure connection and at a specific level of security. T his can be a way to make
changes to sensitive information only over a secure connection, like requiring that password changes be
made over T LS or SASL:
aci: (targetattr="userPassword")(version 3.0; acl "Require secure password
changes"; allow (write) userdn="ldap:///self" and ssf>="56";)
12.9.11. Defining Permissions for DNs That Contain a Comma
DNs that contain commas require special treatment within your LDIF ACI statements. In the target and
bind rule portions of the ACI statement, commas must be escaped by a single backslash (\). For
example:
dn: dc=example.com Bolivia\, S.A.,dc=com
objectClass: top
objectClass: organization
aci: (target="ldap:///dc=example.com Bolivia\,S.A.,dc=com")(targetattr=*)
(version 3.0; acl "aci 2"; allow (all)
groupdn = "ldap:///cn=Directory Administrators,dc=example.com Bolivia\,
S.A.,dc=com";)
510
Chapter 12. Managing Access Control
12.9.12. Proxied Authorization ACI Example
Proxied authorization allows one user to bind and perform operation as another user. For example,
Example Corp. has an accounting program which must be able to bind to the directory as an accounting
administrator in order to write data. T his authorization assumes three things:
T he client application's bind DN is
"uid=MoneyWizAcctSoftware,ou=Applications,dc=exam ple,dc=com ".
T he targeted subtree to which the client application is requesting access is
ou=Accounting,dc=exam ple,dc=com .
An accounting administrator with access permissions to the
ou=Accounting,dc=exam ple,dc=com subtree exists in the directory.
In order for the client application to gain access to the accounting subtree, using the same access
permissions as the accounting administrator, two ACIs must be set:
T he accounting administrator must have access permissions to the
ou=Accounting,dc=exam ple,dc=com subtree, so the following ACI grants all rights to the
accounting administrator entry:
aci: (target="ldap:///ou=Accounting,dc=example,dc=com") (targetattr="*")
(version 3.0; acl "allowAll-AcctAdmin"; allow (all)
userdn="ldap://uid=AcctAdministrator,ou=Administrators,dc=example,dc=com")
T here must be an ACI granting proxy rights to the client application in the directory:
aci: (target="ldap:///ou=Accounting,dc=example,dc=com") (targetattr="*")
(version 3.0; acl "allow proxy-accounting software"; allow (proxy)
userdn="ldap://uid=MoneyWizAcctSoftware,ou=Applications,dc=example,dc=com")
With this ACI in place, the MoneyWizAcctSoftware client application can bind to the directory and send
an LDAP command such as ldapsearch or ldapm odify that requires the access rights of the proxy
DN.
If the client performs an ldapsearch command, the command must include the -Y option for the proxy
account:
/usr/lib64/mozldap/ldapsearch -D
"uid=MoneyWizAcctSoftware,ou=Applications,dc=example,dc=com" -w secret -Y
"dn:uid=AcctAdministrator,ou=Administrators,dc=example,dc=com" ...
T he client or application (MoneyWizAcctSoftware) binds as itself but is granted the privileges of the
proxy entry (AcctAdm inistrator). T he client does not need the password of the proxy entry.
NOTE
T here are some restrictions on binding with proxy authorization. You cannot use the Directory
Manager's DN (root DN) as a proxy DN. Additionally, if Directory Server receives more than one
proxied authentication control, an error is returned to the client application, and the bind attempt is
unsuccessful.
511
Red Hat D irectory Server 8.2 Administration Guide
12.10. Advanced Access Control: Using Macro ACIs
In organizations that use repeating directory tree structures, it is possible to optimize the number of
ACIs used in the directory by using macros. Reducing the number of ACIs in your directory tree makes it
easier to manage your access control policy and improves the efficiency of ACI memory usage.
Macros are placeholders that are used to represent a DN, or a portion of a DN, in an ACI. You can use a
macro to represent a DN in the target portion of the ACI or in the bind rule portion, or both. In practice,
when Directory Server gets an incoming LDAP operation, the ACI macros are matched against the
resource targeted by the LDAP operation. If there is a match, the macro is replaced by the value of the
DN of the targeted resource. Directory Server then evaluates the ACI normally.
12.10.1. Macro ACI Example
Figure 12.3, “Example Directory T ree for Macro ACIs” shows a directory tree which uses macro ACIs to
effectively reduce the overall number of ACIs. T his illustration uses repeating pattern of subdomains with
the same tree structure (ou=groups, ou=people). T his pattern is also repeated across the tree
because the Example Corp. directory tree stores the suffixes
dc=hostedCom pany2,dc=exam ple,dc=com and dc=hostedCom pany3,dc=exam ple,dc=com .
T he ACIs that apply in the directory tree also have a repeating pattern. For example, the following ACI is
located on the dc=hostedCom pany1,dc=exam ple,dc=com node:
aci: (targetattr="*")(targetfilter=(objectClass=nsManagedDomain))
(version 3.0; acl "Domain access"; allow (read,search)
groupdn="ldap:///cn=DomainAdmins,ou=Groups,dc=hostedCompany1,dc=example,dc=com";)
T his ACI grants read and search rights to the Dom ainAdm ins group to any entry in the
dc=hostedCom pany1,dc=exam ple,dc=com tree.
Figure 12.3. Example Directory T ree for Macro ACIs
512
Chapter 12. Managing Access Control
T he following ACI is located on the dc=hostedCom pany1,dc=exam ple,dc=com node:
aci: (targetattr="*")(targetfilter=(objectClass=nsManagedDomain))
(version 3.0; acl "Domain access"; allow (read,search)
groupdn="ldap:///cn=DomainAdmins,ou=Groups,dc=hostedCompany1,dc=example,dc=com";)
T he following ACI is located on the dc=subdom ain1,dc=hostedCom pany1,dc=exam ple,dc=com
node:
aci: (targetattr="*")(targetfilter=(objectClass=nsManagedDomain))
(version 3.0; acl "Domain access"; allow (read,search)
groupdn="ldap:///cn=DomainAdmins,ou=Groups,dc=subdomain1,dc=hostedCompany1,dc=exa
mple,dc=com";)
T he following ACI is located on the dc=hostedCom pany2,dc=exam ple,dc=com node:
aci: (targetattr="*")(targetfilter=(objectClass=nsManagedDomain))
(version 3.0; acl "Domain access"; allow (read,search)
groupdn="ldap:///cn=DomainAdmins,ou=Groups,dc=hostedCompany2,dc=example,dc=com";)
T he following ACI is located on the dc=subdom ain1,dc=hostedCom pany2,dc=exam ple,dc=com
node:
aci: (targetattr="*")(targetfilter=(objectClass=nsManagedDomain))
(version 3.0; acl "Domain access"; allow (read,search)
groupdn="ldap:///cn=DomainAdmins,ou=Groups,dc=subdomain1,dc=hostedCompany2,dc=exam
ple,dc=com";)
In the four ACIs shown above, the only differentiator is the DN specified in the groupdn keyword. By
using a macro for the DN, it is possible to replace these ACIs by a single ACI at the root of the tree, on
the dc=exam ple,dc=com node. T his ACI reads as follows:
aci: (target="ldap:///ou=Groups,($dn),dc=example,dc=com")
(targetattr="*")(targetfilter=(objectClass=nsManagedDomain))
(version 3.0; acl "Domain access"; allow (read,search)
groupdn="ldap:///cn=DomainAdmins,ou=Groups,[$dn],dc=example,dc=com";)
T he target keyword, which was not previously used, is utilized in the new ACI.
In this example, the number of ACIs is reduced from four to one. T he real benefit is a factor of how many
repeating patterns you have down and across your directory tree.
12.10.2. Macro ACI Syntax
Macro ACIs include the following types of expressions to replace a DN or part of a DN:
($dn)
[$dn]
($attr.attrName), where attrName represents an attribute contained in the target entry
513
Red Hat D irectory Server 8.2 Administration Guide
In this section, the ACI keywords used to provide bind credentials, such as userdn, roledn, groupdn,
and userattr, are collectively called the subject, as opposed to the target, of the ACI. Macro ACIs can
be used in the target part or the subject part of an ACI.
T able 12.9, “Macros in ACI Keywords” shows in what parts of the ACI you can use DN macros:
T able 12.9. Macros in ACI Keywords
Macro
ACI Keyword
($dn)
target, targetfilter, userdn, roledn, groupdn,
userattr
[$dn]
targetfilter, userdn, roledn, groupdn, userattr
($attr.attrName)
userdn, roledn, groupdn, userattr
T he following restrictions apply:
If you use ($dn) in targetfilter, userdn, roledn, groupdn, userattr, you must define a
target that contains ($dn).
If you use [$dn] in targetfilter, userdn, roledn, groupdn, userattr, you must define a
target that contains ($dn).
NOTE
When using any macro, you always need a target definition that contains the ($dn) macro.
You can combine the ($dn) macro and the ($attr.attrName) macro.
12.10.2.1. Macro Matching for ($dn)
T he ($dn) macro is replaced by the matching part of the resource targeted in an LDAP request. For
example, you have an LDAP request targeted at the
cn=all,ou=groups,dc=subdom ain1,dc=hostedCom pany1,dc=exam ple,dc=com entry and
an ACI that defines the target as follows:
(target="ldap:///ou=Groups,($dn),dc=example,dc=com")
T he ($dn) macro matches with dc=subdom ain1,dc=hostedCom pany1.
When the subject of the ACI also uses ($dn), the substring that matches the target is used to expand
the subject. For example:
aci: (target="ldap:///ou=*,($dn),dc=example,dc=com")
(targetattr = "*") (version 3.0; acl "Domain access"; allow (read,search)
groupdn="ldap:///cn=DomainAdmins,ou=Groups,($dn),dc=example,dc=com";)
In this case, if the string matching ($dn) in the target is dc=subdom ain1,dc=hostedCom pany1,
then the same string is used in the subject. T he ACI is then expanded as follows:
514
Chapter 12. Managing Access Control
aci: (target="ldap:///ou=Groups,dc=subdomain1,dc=hostedCompany1,
dc=example,dc=com") (targetattr = "*") (version 3.0; acl "Domain
access"; allow (read,search) groupdn="ldap:///cn=DomainAdmins,ou=Groups,
dc=subdomain1,dc=hostedCompany1,dc=example,dc=com";)
Once the macro has been expanded, Directory Server evaluates the ACI following the normal process to
determine whether access is granted.
12.10.2.2. Macro Matching for [$dn]
T he matching mechanism for [$dn] is slightly different than for ($dn). T he DN of the targeted
resource is examined several times, each time dropping the left-most RDN component, until a match is
found.
For example, you have an LDAP request targeted at the
cn=all,ou=groups,dc=subdom ain1,dc=hostedCom pany1,dc=exam ple,dc=com subtree
and the following ACI:
aci: (target="ldap:///ou=Groups,($dn),dc=example,dc=com")
(targetattr = "*") (version 3.0; acl "Domain access"; allow (read,search)
groupdn="ldap:///cn=DomainAdmins,ou=Groups,[$dn],dc=example,dc=com";)
T he steps for expanding this ACI are as follows:
1. ($dn) in the target matches dc=subdom ain1,dc=hostedCom pany1.
2. [$dn] in the subject is replaces with dc=subdom ain1,dc=hostedCom pany1.
T he result is
groupdn="ldap:///cn=Dom ainAdm ins,ou=Groups,dc=subdom ain1,dc=hostedCom p
any1,dc=exam ple,dc=com ". If the bind DN is a member of that group, the matching process
stops, and the ACI is evaluated. If it does not match, the process continues.
3. [$dn] in the subject is replaced with dc=hostedCom pany1.
T he result is
groupdn="ldap:///cn=Dom ainAdm ins,ou=Groups,dc=hostedCom pany1,dc=exam pl
e,dc=com ". In this case, if the bind DN is not a member of that group, the ACI is not evaluated. If
it is a member, the ACI is evaluated.
T he advantage of the [$dn] macro is that it provides a flexible way of granting access to domain-level
administrators to all the subdomains in the directory tree. T herefore, it is useful for expressing a
hierarchical relationship between domains.
For example, consider the following ACI:
aci: (target="ldap:///ou=*, ($dn),dc=example,dc=com")
(targetattr="*")(targetfilter=(objectClass=nsManagedDomain))
(version 3.0; acl "Domain access"; allow (read,search)
groupdn="ldap:///cn=DomainAdmins,ou=Groups,[$dn],dc=example,dc=com";)
It grants access to the members of
cn=Dom ainAdm ins,ou=Groups,dc=hostedCom pany1,dc=exam ple,dc=com to all of the
subdomains under dc=hostedCom pany1, so an administrator belonging to that group could access a
subtree like ou=people,dc=subdom ain1.1,dc=subdom ain1.
However, at the same time, members of cn=Dom ainAdm ins,ou=Groups,dc=subdom ain1.1 would
be denied access to the ou=people,dc=hostedCom pany1 and
515
Red Hat D irectory Server 8.2 Administration Guide
ou=people,dc=hostedCom pany1 nodes.
12.10.2.3. Macro Matching for ($attr.attrName)
T he ($attr.attrName) macro is always used in the subject part of a DN. For example, define the
following roledn:
roledn = "ldap:///cn=DomainAdmins,($attr.ou)"
Now, assume the server receives an LDAP operation targeted at the following entry:
dn:
cn:
sn:
ou:
...
cn=Jane Doe,ou=People,dc=HostedCompany1,dc=example,dc=com
Jane Doe
Doe
Engineering,dc=HostedCompany1,dc=example,dc=com
In order to evaluate the roledn part of the ACI, the server looks at the ou attribute stored in the
targeted entry and uses the value of this attribute to expand the macro. T herefore, in the example, the
roledn is expanded as follows:
roledn =
"ldap:///cn=DomainAdmins,ou=Engineering,dc=HostedCompany1,dc=example,dc=com"
T he Directory Server then evaluates the ACI according to the normal ACI evaluation algorithm.
When an attribute is multi-valued, each value is used to expand the macro, and the first one that
provides a successful match is used. For example:
dn:
cn:
sn:
ou:
ou:
cn=Jane Doe,ou=People,dc=HostedCompany1,dc=example,dc=com
Jane Doe
Doe
Engineering,dc=HostedCompany1,dc=example,dc=com
People,dc=HostedCompany1,dc=example,dc=com...
In this case, when the Directory Server evaluates the ACI, it performs a logical OR on the following
expanded expressions:
roledn =
"ldap:///cn=DomainAdmins,ou=Engineering,dc=HostedCompany1,dc=example,dc=com"
roledn = "ldap:///cn=DomainAdmins,ou=People,dc=HostedCompany1,dc=example,dc=com"
12.11. Access Control and Replication
ACIs are stored as attributes of entries;therefore, if an entry containing ACIs is part of a replicated
database, the ACIs are replicated like any other attribute.
ACIs are always evaluated on the Directory Server that services the incoming LDAP requests. T his
means that when a consumer server receives an update request, it returns a referral to the supplier
server before evaluating whether the request can be serviced on the supplier.
12.12. Compatibility with Earlier Releases
516
Chapter 12. Managing Access Control
Some ACI keywords that were used in earlier releases of Directory Server have been deprecated.
However, for reasons of backward compatibility, the following keywords are still supported:
userdnattr
groupdnattr
T herefore, if you have set up a replication agreement between a legacy supplier server and a version
8.2 consumer, there should not be any problems in the replication of ACIs.
[7] The LDAP to o ls referenc ed in this g uid e are Mo z illa LDAP, ins talled with Direc to ry Server in the /usr/l i b 6 4 /m o zl d ap
d irec to ry o n Red Hat Enterp ris e Linux 5 (6 4-b it); d irec to ries fo r o ther p latfo rms are lis ted in Sec tio n 1.3, “ LDAP To o l Lo c atio ns ” .
Ho wever, Red Hat Enterp ris e Linux s ys tems als o inc lud e LDAP to o ls fro m O p enLDAP. It is p o s s ib le to us e the O p enLDAP
c o mmand s as s ho wn in the examp les , b ut yo u mus t us e the - x arg ument to d is ab le SASL and allo w s imp le authentic atio n.
517
Red Hat D irectory Server 8.2 Administration Guide
Chapter 13. Managing User Authentication
When a user connects to the Red Hat Directory Server, first the user is authenticated. T hen, the
directory grants access rights and resource limits to the user depending upon the identity established
during authentication.
T his chapter describes tasks for managing users, including configuring the password and account
lockout policy for the directory, denying groups of users access to the directory, and limiting system
resources available to users depending upon their bind DNs.
13.1. Managing the Password Policy
A password policy minimizes the risks of using passwords by enforcing the following:
Users must change their passwords according to a schedule.
Users must provide non-trivial passwords.
T he password syntax must meet certain complexity requirements.
After establishing a password policy, which can be for the entire directory or for specific subtrees or
users, user passwords can be protected from potential threats by configuring an account lockout policy.
Account lockout protects against hackers who try to break into the directory by repeatedly guessing a
user's password.
For an overview on password policy, see "Designing a Password Policy" in chapter 7, "Designing a
Secure Directory," in the Deployment Guide.
T his section provides information about configuring password and account lockout policies:
Section 13.1.1, “Configuring the Password Policy”
Section 13.1.2, “Setting User Passwords”
Section 13.1.3, “Password Change Extended Operation”
Section 13.2, “Configuring the Account Lockout Policy”
Section 13.2.3.1, “Managing the Account Lockouts and Replication”
Section 13.3, “Synchronizing Passwords”
13.1.1. Configuring the Password Policy
Directory Server supports fine-grained password policy, so password policies can be applied to the
entire directory (global password policy), a particular subtree (subtree level or local password policy), or
a particular user (user level or local password policy).
Essentially, the password policy is comprised of the following information:
The type or level of password policy checks. T his information indicates whether the server should
check for and enforce a global password policy or local (subtree/user level) password policies.
Password add and modify information. T he password information includes password syntax and
password history details.
Bind information. T he bind information includes the number of grace logins permitted, password
aging attributes, and tracking bind failures.
T he sections that follow describe the procedures for configuring the password policy:
Section 13.1.1.1, “Configuring a Global Password Policy Using the Console”
Section 13.1.1.2, “Configuring a Subtree/User Password Policy Using the Console”
518
Chapter 13. Managing User Authentication
Section 13.1.1.3, “Configuring a Global Password Policy Using the Command Line”
Section 13.1.1.4, “Configuring Subtree/User Password Policy Using the Command Line”
TIP
After configuring the password policy, configure an account lockout policy. For details, see
Section 13.2, “Configuring the Account Lockout Policy”.
13.1.1.1. Configuring a Global Password Policy Using the Console
A global password policy applies to every entry in the entire directory.
1. Select the Configuration tab and then the Data node.
2. In the right pane, select the Passwords tab.
T his tab contains the password policy for the entire Directory Server.
3. Set the password policies for how users can change their own passwords.
T o require users to change their password the first time they log on, select the User m ust
change password after reset checkbox.
NOTE
If users are required to reset their password, only the Directory Manager is authorized
to reset the user's password. A regular administrative user cannot force the users to
update their password.
T o allow users to change their own passwords, select the User m ay change password
checkbox.
T o prevent users from changing their password for a specific duration, enter the number of
days in the Allow changes in X day(s) text box. T his keeps users from quickly cycling
through passwords to reuse a password in their password history.
For the server to maintain a history list of passwords used by each user, select the Keep
519
Red Hat D irectory Server 8.2 Administration Guide
password history checkbox. Enter the number of passwords for the server to keep for
each user in the Rem em ber X passwords text box.
4. Set the policies for when passwords expire.
If user passwords should not expire, select the Password never expires radio button.
T o require users to change their passwords periodically, select the Password expires
after X days radio button, and then enter the number of days that a user password is valid.
T he maximum value for the password age is derived by subtracting January 18, 2038, from
today's date. T he entered value must not be set to the maximum value or too close to the
maximum value. Setting the value to the maximum value can cause the Directory Server to fail
to start because the number of seconds will go past the epoch date. In such an event, the
error log will indicate that the password maximum age is invalid. T o resolve this problem,
correct the passwordMaxAge attribute value in the dse.ldif file.
A common policy is to have passwords expire every 30 to 90 days. By default, the password
maximum age is set to 864 0000 seconds (100 days).
If the Password expire after X days radio button is selected, specify how long before
the password expires to send a warning to the user. In the Send Warning X Days Before
Password Expires text enter the number of days before password expiration to send a
warning.
NOTE
It is not necessary to configure the Directory Server to send a warning to users. T he
Directory Server automatically issues a warning the next time the user attempts to log
into the Directory Server Console that the password will soon expire or has expired.
T his is analogous to an operating system warning that reads "Warning: password
will expire in 7 days" when a user logs in.
5. For the server to check the syntax of a user password to make sure it meets the minimum
requirements set by the password policy, select the Check Password Syntax checkbox. T hen,
specify required password complexity, such as the minimum length and required number of
numeric and special characters. T he password syntax requirements are described more in
T able 13.1, “Password Policy Attributes”.
520
Chapter 13. Managing User Authentication
6. From the Password Encryption pull-down menu, select the encryption method for the server
to use when storing passwords.
For detailed information about the encryption methods, refer to the passwordStorageScheme
attribute in T able 13.1, “Password Policy Attributes”.
T he Password Encryption menu might contain other encryption methods, as the directory
dynamically creates the menu depending upon the existing encryption methods it finds in the
directory.
7. Click Save.
13.1.1.2. Configuring a Subtree/User Password Policy Using the Console
1. Enable a fine-grained password policy globally, as described in Section 13.1.1.1, “Configuring a
Global Password Policy Using the Console”. Be sure to check the Enable fine-grained
password policy checkbox to allow user-level password policies.
521
Red Hat D irectory Server 8.2 Administration Guide
NOTE
T he password policy must be enabled globally before it will be applied locally. No other
global password policy features must be set, and the global password policy will not
override the local policy if they differ.
2. Create the local password policy for the subtree or user.
a. Select the Directory tab.
b. In the navigation pane, select the subtree or user entry for which to set up the password
policy.
c. From the Object menu, select the Manage Password Policy option, and then select
the For user or For subtree.
d. In the Passwords tab, select the Create subtree/user level password policy
checkbox to add the required attributes. T he password policy settings — resetting,
expiration, syntax, and encryption — are the same as for the global policy in
Section 13.1.1.1, “Configuring a Global Password Policy Using the Console”.
e. In the Account Lockout tab, specify the appropriate information, and click Save.
522
Chapter 13. Managing User Authentication
13.1.1.3. Configuring a Global Password Policy Using the Command Line
T o set up the password policy for a subtree or user, add the required entries and attributes at the
subtree or user level, set the appropriate values to the password policy attributes, and enable finegrained password policy checking.
No password policy attributes are set by default. Each password policy attribute must be added
manually to the cn=config entry to create a global policy. T hese can be passed all together by
passing an LDIF file with ldapm odify.
1. Create the LDIF file. Each statement is the same as inputting the changes through stdin, with
separate update statements separated by a dash (-).
dn: cn=config
changetype: modify
add: passwordChange
passwordChange: on
add: passwordExp
passwordExp: on
add: passwordMaxAge
passwordMaxAge: 8640000
add: passwordCheckSyntax
passwordCheckSyntax: on
add: passwordMinCategories
passwordMinCategories: 3
add: passwordStorageScheme
passwordStorageScheme: SHA-512
^D
2. Pass the LDIF file to the server using the -f option with the ldapm odify command.
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -f
user-pwdpolicy.ldif
523
Red Hat D irectory Server 8.2 Administration Guide
T able 13.1, “Password Policy Attributes” describes the attributes available to configure the password
policy.
524
Chapter 13. Managing User Authentication
T able 13.1. Password Policy Attributes
Attribute Name
Definition
passwordGraceLimit
T his attribute indicates the number of grace
logins permitted when a user's password is
expired. When set to a positive number, the user
will be allowed to bind with the expired password
for that many times. For the global password
policy, the attribute is defined under cn=config.
By default, this attribute is set to 0, which means
grace logins are not permitted.
passwordMustChange
When on, this attribute requires users to change
their passwords when they first login to the
directory or after the password is reset by the
Directory Manager. T he user is required to
change their password even if user-defined
passwords are disabled. If this attribute is set to
off, passwords assigned by the Directory
Manager should not follow any obvious
convention and should be difficult to discover.
T his attribute is off by default.
passwordChange
When on, this attribute indicates that users may
change their own password. Allowing users to set
their own passwords runs the risk of users
choosing passwords that are easy to remember.
However, setting good passwords for the user
requires a significant administrative effort. In
addition, providing passwords to users that are
not meaningful to them runs the risk that users
will write the password down somewhere that can
be discovered. T his attribute is on by default.
passwordExp
When on, this attribute indicates that the user's
password will expire after an interval given by the
passwordMaxAge attribute. Making passwords
expire helps protect the directory data because
the longer a password is in use, the more likely it
is to be discovered. T his attribute is off by
default.
passwordMaxAge
T his attribute indicates the number of seconds
after which user passwords expire. T o use this
attribute, enable password expiration using the
passwordExp attribute. T his attribute is a
dynamic parameter in that its maximum value is
derived by subtracting January 18, 2038, from
today's date. T he attribute value must not be set
to the maximum value or too close to the
maximum value. If the value is set to the maximum
value, Directory Server may fail to start because
the number of seconds will go past the epoch
date. In such an event, the error log will indicate
that the password maximum age is invalid. T o
resolve this problem, correct the
525
Red Hat D irectory Server 8.2 Administration Guide
resolve this problem, correct the
passwordMaxAge attribute value in the
dse.ldif file. A common policy is to have
passwords expire every 30 to 90 days. By
default, the password maximum age is set to
864 0000 seconds (100 days).
passwordWarning
T his attribute indicates the number of seconds
before a warning message is sent to users
whose password is about to expire. Depending
on the LDAP client application, users may be
prompted to change their password when the
warning is sent. By default, the directory sends
the warning 864 00 seconds (1 day) before the
password is about to expire. However, a
password never expires until the warning
message has been sent. T herefore, if users don't
bind to the Directory Server for longer than the
passwordMaxAge, they will still get the warning
message in time to change their password.
passwordMinAge
T his attribute indicates the number of seconds
that must pass before a user can change their
password. Use this attribute in conjunction with
the passwordInHistory attribute to discourage
users from reusing old passwords. For example,
setting the minimum password age to 2 days
prevents users from repeatedly changing their
passwords during a single session to cycle
through the password history and reuse an old
password once it has been removed from the
history list. T he minimum age can be from 0 to
214 74 72000 seconds (24,855 days). A value of
zero indicates that the user can change the
password immediately. T he default value of this
attribute is 0.
passwordHistory
T his attribute indicates whether the directory
stores a password history. When set to on, the
directory stores the number of passwords
specified in the passwordInHistory attribute in
a history. If a user attempts to reuse one of the
passwords, the password will be rejected. When
this attribute is set to off, any passwords stored
in the history remain there. When this attribute is
set back to on, users will not be able to reuse the
passwords recorded in the history before the
attribute was disabled. T his attribute is off by
default, meaning users can reuse old passwords.
passwordInHistory
T his attribute indicates the number of passwords
the directory stores in the history. T here can be 2
to 24 passwords stored in the history. T his
feature is not enabled unless the
passwordHistory attribute is set to on. T his
attribute is set to 6 by default.
526
Chapter 13. Managing User Authentication
passwordCheckSyntax
When on, this attribute indicates that the
password syntax is checked by the server before
the password is saved. Password syntax
checking ensures that the password string meets
or exceeds the length and complexity
requirements and that the string does not contain
any trivial words. A trivial word is any value stored
in the uid, cn, sn, givenname, ou, or mail
attributes of the user's entry. T his attribute is
off by default.
passwordMinLength
T his attribute specifies the minimum number of
characters that must be used in passwords.
Shorter passwords are easier to crack.
Passwords can be two (2) to 512 characters long.
Generally, a length of eight characters is long
enough to be difficult to crack but short enough
for users to remember without writing it down.
T his attribute is set to 8 by default.
passwordMaxRepeats
T his attribute set the maximum number of times
that the same character can be used in row, such
as aaaaa. Setting the attribute to 0 means that
there is no limit on how many time a character can
be repeated. T his attribute is set to 0 by default.
passwordMinAlphas
T his attribute sets the minimum number of
alphabetic characters that must be used in the
password. T his setting does not set any
requirements for the letter case; requirements for
the minimum number of lowercase and uppercase
letters are set in the passwordMinLower and
passwordMinUpper attributes, respectively. By
default, this attribute is set to 0, meaning there is
no required minimum.
passwordMinDigits
T his attribute sets the minimum number of
numeric characters (0 through 9) which must be
used in the password. By default, this attribute is
set to 0, meaning there is no required minimum.
passwordMinSpecials
T his attribute sets the minimum number of special
ASCII characters, such as !@ #$., which must be
used in the password. By default, this attribute is
set to 0, meaning there is no required minimum.
passwordMinLowers
T his attribute sets the minimum number of lower
case alphabetic characters, a to z, which must be
used in the password. By default, this attribute is
set to 0, meaning there is no required minimum.
passwordMinCategories
T his attribute sets the minimum number of
categories which must be used in the password.
T here are eight categories available:
Uppercase letters (A to Z )
Lowercase letters (a to z)
527
Red Hat D irectory Server 8.2 Administration Guide
Numbers (0 through 9)
Special ASCII characters, such as $
ASCII alphabetic characters, regardless of case
(a to z and A to Z )
8-bit characters
Repeated characters, such as aaaaaa
T his attribute is set to 3 by default.
passwordMinUppers
T his attribute sets the minimum number of upper
case alphabetic characters, A to Z , which must be
used in the password. By default, this attribute is
set to 0, meaning there is no required minimum.
passwordT okenLength
T his attribute sets the minimum length for any
tokens used with Directory Server. T he token
length can be from 1 to 64 characters. T his
attribute is set to 3 by default.
passwordMin8bit
T his attribute sets the minimum number of 8-bit
characters used in the password. T he default
number is 0, meaning none are required.
passwordStorageScheme
T his attribute specifies the type of encryption
used to store Directory Server passwords. T he
following encryption types are supported by
Directory Server:
SSHA (Salted Secure Hash Algorithm). T his
method is recommended as it is the most secure.
T he Directory Server supports SSHA, SSHA-256,
SSHA-384 , and SSHA-512. SSHA is the default
method.
SHA (Secure Hash Algorithm). A one-way hash
algorithm; it is supported only for backwards
compatibility with Directory Server 4.x and should
not be used otherwise. T his includes support for
SHA, SHA-256, SHA-384 , and SHA-512
algorithms, which protects against some
insecurities in the SHA-1 algorithm.
MD5. MD5 is not as secure as SSHA but is
available for legacy applications require it.
Salted MD5. T his storage scheme is more secure
than plain MD5 hash, but still less secure than
SSHA. T his storage scheme is not included for
use with new passwords but to help with
migrating user accounts from directories which
support salted MD5.
crypt. T he UNIX crypt algorithm, provided for
compatibility with UNIX passwords.
528
Chapter 13. Managing User Authentication
clear. T his encryption type indicates that the
password will appear in plain text.
T he only password storage scheme that can be
used with SASL DIGEST -MD5 is CLEAR.
Passwords stored using crypt, SHA, or SSHA
formats cannot be used for secure login through
SASL DIGEST -MD5. T o provide a customized
storage scheme, consult Red Hat professional
services.
13.1.1.4 . Configuring Subtree/User Password Policy Using the Command Line
1. Add the required attributes to the subtree or user entries by running the ns-newpwpolicy.pl
script.
T he command syntax for the script is as follows:
ns-newpwpolicy.pl [-D rootDN]
h host] -U userDN -S suffixDN
-w password | -w - | -j filename [-p port] [-
For updating a subtree entry, use the -S option. For updating a user entry, use the -U option. T he
ns-newpwpolicy.pl script accepts only one user or subtree entry at a time. It can, however,
accept both user and suffix entries at the same time. For details about the script, see the Directory
Server Configuration and Command-Line Tool Reference.
2. T he script adds the required attributes depending on whether the target entry is a subtree or user
entry.
For a subtree (for example, ou=people,dc=exam ple,dc=com ), the following entries are
added:
A container entry (nsPwPolicyContainer)at the subtree level for holding various password
policy-related entries for the subtree and all its children. For example:
dn: cn=nsPwPolicyContainer,ou=people,dc=example,dc=com
objectClass: top
objectClass: nsContainer
cn: nsPwPolicyContainer
T he actual password policy specification entry (nsPwPolicyEntry) for holding all the
password policy attributes that are specific to the subtree. For example:
dn: cn=cn=nsPwPolicyEntry\,ou=people\,dc=example\,dc=com,
cn=nsPwPolicyContainer,ou=people,dc=example,dc=com
objectclass: top
objectclass: extensibleObject
objectclass: ldapsubentry
objectclass: passwordpolicy
T he CoS template entry (nsPwTemplateEntry) that has the pwdpolicysubentry value
pointing to the above (nsPwPolicyEntry) entry. For example:
529
Red Hat D irectory Server 8.2 Administration Guide
dn: cn=cn=nsPwTemplateEntry\,ou=people\,dc=example\,dc=com,
cn=nsPwPolicyContainer,ou=people,dc=example,dc=com
objectclass: top
objectclass: extensibleObject
objectclass: costemplate
objectclass: ldapsubentry
cosPriority: 1
pwdpolicysubentry: cn=cn=nsPwPolicyEntry\,ou=people\,dc=example\,dc=com,
cn=nsPwPolicyContainer,ou=people,dc=example,dc=com
T he CoS specification entry at the subtree level. For example:
dn: cn=newpwdpolicy_cos,ou=people,dc=example,dc=com
objectclass: top
objectclass: LDAPsubentry
objectclass: cosSuperDefinition
objectclass: cosPointerDefinition
cosTemplateDn: cn=cn=nsPwTemplateEntry\,ou=people\,dc=example,dc=com,
cn=nsPwPolicyContainer,ou=people,dc=example,dc=com
cosAttribute: pwdpolicysubentry default operational
For a user (for example, uid=jdoe,ou=people,dc=exam ple,dc=com ), the following entries
are added:
A container entry (nsPwPolicyContainer) at the parent level for holding various password
policy related entries for the user and all its children. For example:
dn: cn=nsPwPolicyContainer,ou=people,dc=example,dc=com
objectClass: top
objectClass: nsContainer
cn: nsPwPolicyContainer
T he actual password policy specification entry (nsPwPolicyEntry) for holding the password
policy attributes that are specific to the user. For example:
dn: cn=cn=nsPwPolicyEntry\,uid=jdoe\,ou=people\,dc=example\,dc=com,
cn=nsPwPolicyContainer,ou=people,dc=example,dc=com
objectclass: top
objectclass: extensibleObject
objectclass: ldapsubentry
objectclass: passwordpolicy
3. Assign the value of the above entry DN to the pwdpolicysubentry attribute of the target entry.
For example, this assigns the password policy to the user entry:
dn: uid=jdoe,ou=people,dc=example,dc=com
changetype: modify
replace: pwdpolicysubentry
pwdpolicysubentry:
cn=cn=nsPwPolicyEntry\,uid=jdoe\,ou=people\,dc=example\,dc=com,
cn=nsPwPolicyContainer,ou=people,dc=example,dc=com
4. Set the password policy attributes for the subtree or user entry with the appropriate values.
T able 13.1, “Password Policy Attributes” describes the attributes available to configure the
password policy. T he ldapm odify utility can be used to change these attributes in the subtree
or user entry which contains the nsPwPolicyEntry object class.
530
Chapter 13. Managing User Authentication
NOTE
T he nsslapd-pwpolicy-local attribute of the cn=config entry controls the type of
password policy the server enforces. By default, this attribute is disabled (off). When the
attribute is disabled, the server only checks for and enforces the global password policy;
the subtree and user level password policies are ignored. When the nsnewpwpolicy.pl script runs, it first checks for the specified subtree and user entries
and, if they exist, modifies them. After updating the entries successfully, the script sets the
nsslapd-pwpolicy-local configuration parameter to on. If the subtree and user level
password policy should not be enabled, be sure to set nsslapd-pwpolicy-local to off
after running the script.
T o turn off user and subtree level password policy checks, set the nsslapd-pwpolicy-local attribute
to off by modifying the cn=config entry. For example:
[8 ]
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=config
changetype: modify
replace: nsslapd-pwpolicy-local: on
nsslapd-pwpolicy-local: off
T his attribute can also be disabled by modifying it directly in the configuration file (dse.ldif).
1. Stop the server.
service dirsrv stop instance
2. Open the dse.ldif file in a text editor.
3. Set the value of nsslapd-pwpolicy-local to off, and save.
nsslapd-pwpolicy-local: off
4. Start the server.
service dirsrv start instance
13.1.1.5. Manually Setting Default Password Syntax Checking for Local Password Policies
T he settings for the global password policy should be in effect for any local password policies, if those
attributes are not explicitly set. If the password policy attribute is not set in either the global or the local
policy, then the default values should be assumed.
However, there is a bug in Directory Server, so that if a password policy attribute is set in the global
password policy but not in the local password policy, then neither the global setting nor the default
settings is enforced by the local password policy. T o work around this, set the password attributes
explicitly in the local password policy.
1. Enable global syntax checking, as in Section 13.1.1.1, “Configuring a Global Password Policy
Using the Console”, and save the policy.
2. Edit the local password policy to contain all password syntax
531
Red Hat D irectory Server 8.2 Administration Guide
3. Enable fine-grained password checking, as in Section 13.1.1.2, “Configuring a Subtree/User
Password Policy Using the Console”, and save the policy.
4. Edit the local password policy to contain all password syntax attributes. Set the values to
something other than the default settings, as listed in the Configuration and Command-Line Tool
Reference.
5. Re-edit the local password policy with the desired values, even if they are the defaults.
13.1.2. Setting User Passwords
An entry can be used to bind to the directory only if it has a userPassword attribute and if it has not
been inactivated. Because user passwords are stored in the directory, the user passwords can be set
or reset with any LDAP operation, like ldapm odify.[8 ]
For information on creating and modifying directory entries, see Chapter 3, Creating Directory Entries.
For information on inactivating user accounts, refer to Section 13.8, “Inactivating Users and Roles”.
Passwords can also be set and reset in the Users and Groups area of the Red Hat Admin Server or
Directory Server Console. For information on how to use the Users and Groups area in the Admin
Server Console, see the online help that is available in the Red Hat Admin Server.
13.1.3. Password Change Extended Operation
While most passwords can be changed through the Console and other Directory Server features or
through the ldapm odify operation, there are some passwords that cannot be changed through
regular LDAP operations. T hese passwords may be stored outside the Directory Server, such as
passwords stored in a SASL application. T hese passwords can be modified through the password
change extended operation.
Directory Server supports the password change extended operation as defined in RFC 3062, so users
can change their passwords, using a suitable client, in a standards-compliant way. Directory Server
does not include a client application for the password change extended operation. However, the
ldappasswd utility can be used as follows:
ldappasswd -h hostname -p secure_port -Z -P /path/to/cert8.db -D bindDN -w
bindPassword
[-a oldPassword] -s newPassworduser
532
Chapter 13. Managing User Authentication
T able 13.2. ldappasswd Options
Parameter
Description
-h
Gives the hostname of the Directory Server.
-p
Gives the port number of the Directory Server.
Since SSL is required for password change
operations, this is usually give the T LS/SSL port
of the Directory Server. With the -ZZ or -ZZZ for
Start T LS, this can be the standard port.
-Z
Requires SSL for the connection. A secure
connection is required for the password change
operation.
NOTE
ldappasswd also supports Start T LS
encryption (-ZZ[Z]).
-P
Gives the full path to the certificate database
which contains the CA certificate of the CA that
issued the Directory Server client certificate. If the
ldappasswd command is run on the same
machine that the Directory Server is installed on,
this can be
/etc/dirsrv/slapd-instance_name/cert8
.db. If this is not given, the default is the current
directory.
-D
Gives the bind DN.
-w
Gives the password for the bind DN.
-a
Optional. Gives the old password, which is being
changed.
-s
Sets the new password.
T o use Start T LS, which runs the command on a non-secure port, run ldappasswd with the -ZZ option
and the standard LDAP port number. T he password extended change operation has the following
format:
ldappasswd -h hostname -p standard_port -ZZ -P /path/to/cert8.db -D bindDN -w
bindPassword
-s newPassworduser [-a oldPassword]
Use the -ZZZ for additional certificate hostname validation.
T o modify an entry's password, run ldappasswd like any other LDAP operation. It is not necessary to
specify a user if the account is the same as that given in the bind DN. For example:
ldappasswd -h ldap.example.com -p 389 -ZZ -D
"uid=jsmith,ou=People,dc=example,dc=com" -w secret -s newpassword
T o change the password on an entry other than the one specified in the bind credentials, run
533
Red Hat D irectory Server 8.2 Administration Guide
ldappasswd as shown below, adding the user DN to the operation and providing separate credentials,
as follows:
ldappasswd -h server.example.com -p 389 -ZZ -D "cn=Directory Manager" -w secret s newpassword "uid=jsmith,ou=People,dc=example,dc=com"
Access control is enforced for the password change operation. If the bind DN does not have rights to
change the specified password, the operation will fail with an Insufficient rights error.
13.2. Configuring the Account Lockout Policy
T he lockout policy works in conjunction with the password policy to provide further security. T he account
lockout feature protects against hackers who try to break into the directory by repeatedly trying to guess
a user's password. T he password policy can be set so that a specific user is locked out of the directory
after a given number of failed attempts to bind.
Configuring the account lockout policy is described in the following sections:
Section 13.2.1, “Configuring the Account Lockout Policy Using the Console”
Section 13.2.2, “Configuring the Account Lockout Policy Using the Command Line”
13.2.1. Configuring the Account Lockout Policy Using the Console
T o set up or modify the account lockout policy for the Directory Server:
1. Select the Configuration tab and then the Data node.
2. In the right pane, select the Account Lockout tab.
3. T o enable account lockout, select the Accounts m ay be locked out checkbox.
4. Enter the maximum number of allowed bind failures in the Lockout account after X login
failures text box. T he server locks out users who exceed the limit specified here.
5. In the Reset failure counter after X m inutes text box, enter the number of minutes for
the server to wait before resetting the bind failure counter to zero.
6. Set the interval for users to be locked out of the directory.
Select the Lockout Forever radio button to lock users out until their passwords have been
reset by the administrator.
Set a specific lockout period by selecting the Lockout Duration radio button and entering
534
Chapter 13. Managing User Authentication
the time (in minutes) in the text box.
7. Click Save.
13.2.2. Configuring the Account Lockout Policy Using the Command Line
T his section describes the attributes to create an account lockout policy to protect the passwords
stored in the server. Use ldapm odify to change these attributes in the cn=config entry.
T able 13.3, “Account Lockout Policy Attributes” describes the attributes available to configure the
account lockout policy.
535
Red Hat D irectory Server 8.2 Administration Guide
T able 13.3. Account Lockout Policy Attributes
Attribute Name
Definition
passwordLockout
T his attribute indicates whether users are locked
out of the directory after a given number of failed
bind attempts. Set the number of failed bind
attempts after which the user will be locked out
using the passwordMaxFailure attribute. Users
can be locked out for a specific time or until an
administrator resets the password. T his attribute
is set to off by default, meaning that users will
not be locked out of the directory.
passwordMaxFailure
T his attribute indicates the number of failed bind
attempts after which a user will be locked out of
the directory. T his attribute takes affect only if the
passwordLockout attribute is set to on. T his
attribute is set to 3 bind failures by default.
passwordUnlock
T his attribute sets whether a user can log back
into the server without administrator intervation.
T he default is for this attribute to be on, meaning
that the user can log back into the server after a
certain lockout period has passed. If this attribute
is turned off, then the user cannot log back in
using that account until it is manually unlocked by
an administrator.
passwordLockoutDuration
T his attribute indicates the time, in seconds, that
users will be locked out of the directory. T he
passwordUnlock attribute specifies if a user is
locked out until the password is reset by an
administrator (which means that the user is
locked out indefinitely). If the passwordUnlock
attribute is set to on, then the use can log in
again as soon as the lockout duration time is
reached. By default, the user is locked out for
3600 seconds.
passwordResetFailureCount
T his attribute specifies the time, in seconds, after
which the password failure counter will be reset.
Each time an invalid password is sent from the
user's account, the password failure counter is
incremented. If the passwordLockout attribute is
set to on, users will be locked out of the directory
when the counter reaches the number of failures
specified by the passwordMaxFailure attribute.
T he account is locked out for the interval
specified in the passwordLockoutDuration
attribute, after which time the failure counter is
reset to zero (0). Because the counter's purpose
is to gage when a hacker is trying to gain access
to the system, the counter must continue for a
period long enough to detect a hacker. However,
if the counter were to increment indefinitely over
536
Chapter 13. Managing User Authentication
days and weeks, valid users might be locked out
inadvertently. T he reset password failure count
attribute is set 600 seconds by default.
13.2.3. Replicating Account Lockout Attributes
Account lockout policies will block a user ID from being able to access the Directory Server if the login
attempt fails a set number of times. T his prevents hackers or other malicious people from illegitimately
accessing the Directory Server by guessing a password. Password policies are set locally, and
generally account lockout attributes are local to each replica. T his means that a person can attempt to
log in to one replica until the account lockout count is reached, then try again immediately on another
replica. T he way to prevent that is to replicate the attributes related to the account lockout counts for an
entry, so that the malicious user is locked out of every supplier and consumer replica in the configuration
if a login attempt fails on a single master.
By default, three password policy attributes are not replicated, even if other password attributes are.
T hese attributes are related to of login failures and lockout periods:
passwordRetryCount
retryCountResetTime
accountUnlockTime
13.2.3.1. Managing the Account Lockouts and Replication
Password and account lockout policies are enforced in a replicated environment slightly differently:
Password policies are enforced on the data master.
Account lockout is enforced on all servers participating in replication.
Some of the password policy information in the directory is replicated automatically:
passwordMinAge and passwordMaxAge
passwordExp
passwordWarning
However, the configuration information is kept locally and is not replicated. T his information includes the
password syntax and the history of password modifications. Account lockout counters and tiers are not
replicated, either, unless specifically configured for replication.
When configuring a password policy in a replicated environment, make sure that these elements are in
place, so password policies and account lockout settings are enforced consistently:
Warnings from the server of an impending password expiration are issued by all replicas. T his
information is kept locally on each server, so if a user binds to several replicas in turn, they will be
issued the same warning several times. In addition, if the user changes the password, it may take
time for this information to filter to the replicas. If a user changes a password and then immediately
rebinds, he may find that the bind fails until the replica registers the changes.
T he same bind behavior should occur on all servers, including suppliers and replicas. Make sure to
create the same password policy configuration information on each server.
Account lockout counters may not work as expected in a multi-mastered environment. Account
lockout counters are not replicated by default (although this can be configured). If account lockout
attributes are not replicated at all, then a user could be locked out from one server but could
successfully bind to another server (or, conversely, a user may be unlocked on one server and still
537
Red Hat D irectory Server 8.2 Administration Guide
blocked on another). If account lockout attributes are replicated, then there could be lags between an
account lockout change on one server and when that change is propagated to the other servers. It
depends on the replication schedule.
Entries that are created for replication (for example, the server identities) need to have passwords
that never expire. T o make sure that these special users have passwords that do not expire, add the
passwordExpirationTime attribute to the entry, and give it a value of 203801190314 07Z (the top
of the valid range).
13.2.3.2. Configuring Directory Server to Replicate Password Policy Attributes
A special core configuration attribute controls whether password policy operational attributes are
replicated. T his is the passwordIsGlobalPolicy attribute, which is enabled in the consumer Directory
Server configuration to allow the consumer to accept password policy operational attributes.
By default, this attribute is set to off.
T o enable these attributes to be replicated, change the passwordIsGlobalPolicy configuration
attribute on the consumer:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
consumer1.example.com
dn: cn=config
changetype: modify
replace: passwordIsGlobalPolicy
passwordIsGlobalPolicy: on
Changing that value to on allows the passwordRetryCount, retryCountResetTime, and
accountUnlockTime to be replicated. No other configuration is necessary for the attributes to be
included with the replicated attributes.
13.2.3.3. Configuring Fractional Replication for Password Policy Attributes
Setting the passwordIsGlobalPolicy attribute affects the consumer in replication, in that it allows the
consumer to receive updates to those attributes. T o control whether the password policy attributes are
actually replicated by the supplier, use fractional replication, which controls what specific entry attributes
are replicated.
If the password policy attributes should be replicated, then make sure these attributes are included in
the fractional replication agreement (as they are by default).
If the passwordIsGlobalPolicy attribute is set to off on the consumer, so no password policy
attributes should be replicated, use fractional replication (described in Section 9.1.7, “Replicating a
Subset of Attributes with Fractional Replication”) to enforce that on the supplier and specifically exclude
those attributes from the replication agreement.
1. When configuring the replication agreement on the supplier, as described (for example) in
Section 9.4.3, “Creating the Replication Agreement”, select the Enable Fractional
Replication checkbox.
2. By default, every attribute is listed in the Replicated Attributes box. Select the
passwordRetryCount, retryCountResetTime, and accountUnlockTime parameters and click
the arrow button to move them into the Do Not Replicate box.
538
Chapter 13. Managing User Authentication
3. Finish configuring the replication agreement.
13.3. Synchronizing Passwords
Password changes in a Directory Server entry can be synchronized to password attributes in Active
Directory entries by using the Password Sync utility.
When passwords are synchronized, password policies are enforced on each sync peer locally. T he
syntax or minimum length requirements on the Directory Server apply when the password is changed in
the Directory Server. When the changed password is synced over to the Windows server, the Windows
password policy is enforced. T he password policies themselves are not synchronized.
Configuration information is kept locally and cannot be synchronized, including the password change
history and the account lockout counters.
When configuring a password policy for synchronization, consider the following points:
T he Password Sync utility must be installed locally on the Windows machine that will be
synchronized with a Directory Server.
Password Sync can only link the Windows machine to a single Directory Server; to sync changes
with multiple Directory Server instances, configure the Directory Server for multi-master replication.
Password expiration warnings and times, failed bind attempts, and other password-related
information is enforced locally per server and is not synchronized between sync peer servers.
T he same bind behavior should occur on all servers. Make sure to create the same or similar
password policies on both Directory Server and Active Directory servers.
Entries that are created for synchronization (for example, the server identities) need to have
passwords that never expire. T o make sure that these special users have passwords that do not
expire, add the passwordExpirationTime attribute to the Directory Server entry, and give it a value
of 203801190314 07Z (the top of the valid range).
539
Red Hat D irectory Server 8.2 Administration Guide
See Chapter 10, Synchronizing Red Hat Directory Server with Microsoft Active Directory for more
information on synchronizing Directory Server and Windows users and passwords.
13.4. Setting Resource Limits Based on the Bind DN
Server limits for search operations are controlled using special operational attribute values on the client
application binding to the directory. You can set the following search operation limits:
Look through limit. Specifies how many entries can be examined for a search operation.
Size limit. Specifies the maximum number of entries the server returns to a client application in
response to a search operation.
Time limit. Specifies the maximum time the server spends processing a search operation.
Idle timeout. Specifies the time a connection to the server can be idle before the connection is
dropped.
T he resource limits set for the client application take precedence over the default resource limits set for
in the global server configuration.
NOTE
T he Directory Manager receives unlimited resources by default.
Section 13.4.1, “Setting Resource Limits Using the Console”
Section 13.4.2, “Setting Resource Limits Using the Command Line”
Section 13.4.3, “Setting Resource Limits on Anonymous Binds”
13.4.1. Setting Resource Limits Using the Console
1. Select the Directory tab.
2. Browse the navigation tree in the left navigation pane, and double-click the user or role for which
to set resource limits.
T he Edit Entry dialog box appears.
3. Click Account in the left pane.
4. Set the resource limits. T here are four different limits that can be set:
Look through limit. T he maximum number of entries are examined for a search operation.
Size limit. T he maximum number of entries the server returns to a client application in
response to a search operation.
Time limit. T he maximum time the server spends processing a search operation.
Idle timeout. T he time a connection to the server can be idle before the connection is dropped.
54 0
Chapter 13. Managing User Authentication
Entering a value of -1 indicates no limit.
5. Click OK.
13.4.2. Setting Resource Limits Using the Command Line
T he following operational attributes can be set for each entry using the command line. Use
ldapm odify to add the following attributes to the entry:
Attribute
Description
nsdslapd-lookthroughlimit
Specifies how many entries are examined for a
search operation. Giving this attribute a value of 1 indicates that there is no limit.
nsslapd-sizelimit
Specifies the maximum number of entries the
server returns to a client application in response
to a search operation. Giving this attribute a value
of -1 indicates that there is no limit.
nsslapd-timelimit
Specifies the maximum time the server spends
processing a search operation. Giving this
attribute a value of -1 indicates that there is no
time limit.
nsidletimeout
Specifies the time a connection to the server can
be idle before the connection is dropped. T he
value is given in seconds. Giving this attribute a
value of -1 indicates that there is no limit.
For example, this sets the size limit for Barbara Jensen by using ldapm odify[8 ] to modify her entry:
54 1
Red Hat D irectory Server 8.2 Administration Guide
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: uid=bjensen,ou=people,dc=example,dc=com
changetype: modify
add:nsslapd-sizelimit
nsslapd-sizelimit: 500
T he ldapm odify statement adds the nsSizeLimit attribute to Babs Jensen's entry and gives it a
search return size limit of 500 entries.
13.4.3. Setting Resource Limits on Anonymous Binds
Resource limits are set on a user entry. An anonymous bind, obviously, doesn't have a user entry
associated with it. T his means that the global resource limits usually apply to anonymous operations.
However, it is possible to configure resource limits specifically for anonymous binds by creating a
template user entry that has resource limits, and then applying that template to anonymous binds.
1. Create a template entry and set whatever resource limits you want to apply to anonymous binds.
TIP
For performance reasons, the template should be in the normal backend, not in the
cn=config suffix, which doesn't use an entry cache.
For example:
/usr/lib64/mozldap/ldapmodify -a -D "cn=directory manager" -w secret -p 389
-h server.example.com
dn: cn=anon template,ou=people,dc=example,dc=com
objectclass: person
objectclass: top
cn: anon template
sn: template
nsSizeLimit: 250
nsLookThroughLimit: 1000
nsTimeLimit: 60
2. Add the nsslapd-anonlimitsdn to the server configuration, pointing to the DN of the template
entry. Any of the resource limits in Section 13.4.2, “Setting Resource Limits Using the Command
Line” can be set. For example:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389
dn: cn=config
changetype: modify
add: nsslapd-anonlimitsdn
nsslapd-anonlimitsdn: cn=anon template,ou=people,dc=example,dc=com
13.5. Enabling Different Types of Binds
Whenever an entity logs into or accesses the Directory Server, it binds to the directory. T here are many
different types of bind operation, sometimes depending on the method of binding (such as simple binds
54 2
Chapter 13. Managing User Authentication
or autobind) and some depending on the identity of user binding to the directory (anonymous and
unauthenticated binds).
T he following sections contain configuration parameters that can increase the security of binds (as in
Section 13.5.1, “Requiring Secure Binds”) or streamline bind operations (such as Section 13.5.4,
“Configuring Autobind”).
13.5.1. Requiring Secure Binds
A simple bind is when an entity uses a simple bind DN-password combination to authenticate to the
Directory Server. Although it is possible to use a password file rather than sending a password directly
through the command line, both methods still require sending or accessing a plaintext password over
the wire. T hat makes the password vulnerable to anyone sniffing the connection.
It is possible to require simple binds to occur over a secure connection (SSL/T LS or Start T LS), which
effectively encrypts the plaintext password as it is sent with the bind operation. (It is also possible to use
alternatives to simple binds, such as SASL authentication and certificate-based authentication.)
IMPORTANT
Along with regular users logging into the server and LDAP operations, server-to-server
connections are affected by requiring secure connections for simple binds. Replication,
synchronization, and database chaining can all use simple binds between servers, for instance.
Make sure that replication and sync agreements and chaining configuration specify secure
connections if the nsslapd-require-secure-binds attribute is turned on. Otherwise, these
operations will fail.
NOTE
Requiring a secure connection for bind operations only applies to authenticated binds. Bind
operations without a password (anonymous and unauthenticated binds) can proceed over
standard connections.
1. Add the nsslapd-require-secure-binds attribute to the cn=config entry:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389
dn: cn=config
changetype: modify
replace: nsslapd-require-secure-binds
nsslapd-require-secure-binds: on
2. Restart the server.
service dirsrv restart
13.5.2. Disabling Anonymous Binds
If a user attempts to connect to the Directory Server without supplying any username or password, this
is an anonymous bind. Anonymous binds simplify common search and read operations, like checking the
directory for a phone number or email address, by not requiring users to authenticate to the directory
first.
54 3
Red Hat D irectory Server 8.2 Administration Guide
However, there are risks with anonymous binds. Adequate ACIs must be in place to restrict access to
sensitive information and to disallow actions like modifies and deletes. Additionally, anonymous binds
can be used for denial of service attacks or for malicious people to gain access to the server.
Section 12.9.1, “Granting Anonymous Access” has examples on setting ACIs to control what anonymous
users can access, and Section 13.4.3, “Setting Resource Limits on Anonymous Binds” has information
on placing resource limits for anonymous users.
If those options do not offer a sufficient level of security, then anonymous binds can be disabled entirely.
1. Add the nsslapd-allow-anonymous-access attribute to the cn=config entry:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389
dn: cn=config
changetype: modify
replace: nsslapd-allow-anonymous-access
nsslapd-allow-anonymous-access: off
2. Restart the server.
service dirsrv restart
NOTE
When anonymous binds are disabled, unauthenticated binds are also disabled automatically.
13.5.3. Allowing Unauthenticated Binds
An unauthenticated bind is a bind where the user supplies a username but not a password. For
example, running an ldapsearch without supplying a password option:
/usr/lib64/mozldap/ldapsearch -D "cn=directory manager" -b "dc=example,dc=com" -s
sub "(objectclass=*)"
When unauthenticated binds are allowed, the bind attempt goes through as an anonymous bind.
Unauthenticated binds are less secure than authenticated binds, and in some directories can be used to
circumvent ACIs or performs denial-of-service attacks. T his is why in Directory Server unauthenticated
binds are disabled by default. If a user tries to bind without a password, the attempt fails:
ldap_simple_bind: DSA is unwilling to perform
ldap_simple_bind: additional info: Unauthenticated binds are not allowed
Unauthenticated binds only apply to bind attempts where a password is not given but a bind identity is. If
the wrong password is given, the operation fails with an invalid credentials error:
ldap_simple_bind: Invalid credentials
If no bind ID or password is given, then the directory returns whatever information is allowed for an
anonymous bind.
T he nsslapd-allow-unauthenticated-binds attribute sets whether to allow an unauthenticated bind
54 4
Chapter 13. Managing User Authentication
to succeed as an anonymous bind. Setting this parameter to on allows unauthenticated binds. By
default, this parameter is off.
1. T o configure unauthenticated binds, edit the Directory Server cn=config entry:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=config
changetype:replace
replace: nsslapd-allow-unauthenticated-binds
nsslapd-allow-unauthenticated-binds: on
2. Restart the server.
service dirsrv restart
13.5.4. Configuring Autobind
Autobind is a way to connect to the Directory Server based on local UNIX credentials, which are mapped
to an identity stored in the directory itself. Autobind is configured in two parts:
Before configuring autobind, first make sure that LDAPI is enabled (in Section 1.6, “Enabling LDAPI”).
T hen, configure the autobind mappings (in Section 13.5.4.2, “Configuring Autobind”).
13.5.4 .1. Overview of Autobind and LDAPI
Inter-process communication (IPC) is a way for separate processes on a Unix machine or a network to
communicate directly with each other. LDAPI is a way to run LDAP connections over these IPC
connections, meaning that LDAP operations can run over Unix sockets. T hese connections are much
faster and more secure than regular LDAP connections.
T he Directory Server uses these LDAPI connections to allow users to bind immediately to the Directory
Server or to access the Directory Server using tools which support connections over Unix sockets.
Autobind uses the uid:gid of the Unix user and maps that user to an entry in the Directory Server, then
allows access for that user.
Autobind allows mappings to three directory entries:
User entries, if the Unix user matches one user entry
Directory Manager (or the super user defined in nsslapd-ldapimaprootdn), if the Unix user is
root
54 5
Red Hat D irectory Server 8.2 Administration Guide
Figure 13.1. Autobind Connection Path
T he special autobind users are entries beneath a special autobind suffix (outside the regular user
subtree). T he entries underneath are identified by their user and group ID numbers:
gidNumber=gid+uidNumberuid, autobindsuffix
If autobind is not enabled but LDAPI is, then Unix users are anonymously bound to the Directory Server,
unless they provide other bind credentials.
NOTE
Autobind allows a client to send a request to the Directory Server without supplying a bind
username and password or using other SASL authentication mechanism. According to the LDAP
standard, if bind information is not given with the request, the server processes the request as an
anonymous bind. T o be compliant with the standard, which requires some kind of bind
information, any client that uses autobind should send the request with SASL/EXT ERNAL.
For more information on configuring SASL, see Section 14.3, “Using SASL”.
13.5.4 .2. Configuring Autobind
Configuring autobind alone allows anonymous access to the Directory Server. It is possible to enable
mapping Unix users to entries and also to map root to Directory Manager.
54 6
Chapter 13. Managing User Authentication
1. Run ldapm odify to update the Directory Server configuration.
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=config
changetype: modify
2. Enable autobind.
replace: nsslapd-ldapiautobind
nsslapd-ldapiautobind: on
3. T o map user entries, add four attributes:
nsslapd-ldapimaptoentries to enable entry mapping
nsslapd-ldapiuidnumbertype to set the Directory Server attribute to map to the Unix UID
number
nsslapd-ldapigidnumbertype to set the Directory Server attribute to map to the Unix group
ID number
nsslapd-ldapientrysearchbase to set the search base to use to find Directory Server user
entries
add: nsslapd-ldapimaptoentries
nsslapd-ldapimaptoentries: on
add: nsslapd-ldapiuidnumbertype
nsslapd-ldapiuidnumbertype: uidNumber
add: nsslapd-ldapigidnumbertype
nsslapd-ldapigidnumbertype: gidNumber
add: nsslapd-ldapientrysearchbase
nsslapd-ldapientrysearchbase: ou=people,dc=example,dc=com
4. T o map the root entry to Directory Manager, add the nsslapd-ldapimaprootdn attribute:
add: nsslapd-ldapimaprootdn
nsslapd-ldapimaprootdn: cn=Directory Manager
5. Restart the server to apply the new configuration.
service dirsrv restart example
13.6. Using Pass-through Authentication
Pass-through authentication (PT A) is a mechanism which allows one Red Hat Directory Server instance
to consult another to authenticate bind requests. Pass-through authentication is implement through the
PT A Plug-in; when enabled, the plug-in lets a Directory Server instance accept simple bind operations
(password-based) for entries not stored in its local database.
Directory Server uses PT A to administer the user and configuration directories on separate instances of
Directory Server.
If the configuration directory and the user directory are installed on separate instances of Directory
54 7
Red Hat D irectory Server 8.2 Administration Guide
Server, the setup program automatically sets up PT A to allow the Configuration Administrator user
(usually adm in) to perform administrative duties.
PT A is required in this case because the adm in user entry is stored under o=NetscapeRoot suffix in
the configuration directory. T herefore, attempts to bind to the user directory as adm in would normally
fail. PT A allows the user directory to transmit the credentials to the configuration directory, which verifies
them. T he user directory then allows the adm in user to bind.
T he user directory in this example acts as the PTA Directory Server, the server that passes through
bind requests to another Directory Server. T he configuration directory acts as the authenticating
directory, the server that contains the entry and verifies the bind credentials of the requesting client.
T he pass-through subtree is the subtree not present on the PT A directory. When a user's bind DN
contains this subtree, the user's credentials are passed on to the authenticating directory.
Figure 13.2. Simple Pass-through Authentication Process
NOTE
T he PT A Plug-in may not be listed in the Directory Server Console if the same server instance is
used for the user directory and the configuration directory.
Here's how pass-through authentication works:
1. T he configuration Directory Server (authenticating directory) is installed on machine A. T he
configuration directory always contains the configuration database and suffix, o=NetscapeRoot.
In this example, the server name is configdir.exam ple.com .
2. T he user Directory Server (PT A directory) is then installed on machine B. T he user directory
stores the root suffix, such as dc=exam ple,dc=com . In this example, the server name is
userdir.exam ple.com .
3. When the user directory is set up on machine B, the setup script prompts for the LDAP URL of the
configuration directory on machine A.
4. T he setup program enables the PT A Plug-in and configures it to use the configuration directory
LDAP URL.
T his entry contains the LDAP URL for the configuration directory. For example:
54 8
Chapter 13. Managing User Authentication
T he user directory is now configured to send all bind requests for entries with a DN containing
o=NetscapeRoot to the configuration directory configdir.exam ple.com .
5. When installation is complete, the adm in user attempts to connect to the user directory to begin
adding users.
6. T he setup program adds the adm in user's entry to the directory as
uid=adm in,ou=T opologyManagem ent,o=NetscapeRoot. So the user directory passes the
bind request through to the configuration directory as defined by the PT A Plug-in configuration.
7. T he configuration directory authenticates the user's credentials and sends the information back to
the user directory.
8. T he user directory allows the adm in user to bind.
13.6.1. PTA Plug-in Syntax
PT A Plug-in configuration information is specified in the cn=Pass T hrough Authentication,
cn=plugins,cn=config entry on the PT A directory (the user directory configured to pass through
bind requests to the authenticating directory) using the required PT A syntax. T here are only two
attributes in this entry that are significant:
nsslapd-pluginEnabled, which sets whether the plug-in is enabled or disabled. T he value for this
attribute can be on or off.
nsslapd-pluginarg0, which points to the configuration directory. T he value for this attribute is the
LDAP URL of the server and suffix to which to pass the bind requests, along with the optional
parameters, maxconns, maxops, timeout, ldver, connlifetime, startTLS.
T he variable components of the PT A plug-in syntax are described in T able 13.4, “PT A Plug-in
Parameters”.
NOTE
T he LDAP URL (ldap|ldaps://authDS/subtree) must be separated from the optional
parameters (maxconns, maxops, timeout, ldver, connlifetime, startTLS) by a single space. If any
of the optional parameters are defined, all of them must be defined, even if only the default values
are used.
Several authenticating directories or subtrees can be specified by incrementing the nsslapdpluginarg attribute suffix by one each time, as in Section 13.6.3.2, “Specifying Multiple
Authenticating Directory Servers”. For example:
nsslapd-pluginarg0: LDAP URL for the first server
nsslapd-pluginarg1: LDAP URL for the second server
nsslapd-pluginarg2: LDAP URL for the third server
...
T he optional parameters are described in the following table in the order in which they appear in the
syntax.
54 9
Red Hat D irectory Server 8.2 Administration Guide
T able 13.4 . PT A Plug-in Parameters
Variable
Definition
state
Defines whether the plug-in is enabled or
disabled. Acceptable values are on or off.
ldap|ldaps
Defines whether SSL is used for communication
between the two Directory Servers. See
Section 13.6.2.1, “Configuring the Servers to Use
a Secure Connection” for more information.
authDS
T he authenticating directory hostname. T he port
number of the Directory Server can be given by
adding a colon and then the port number. For
example,
ldap://dirserver.exam ple.com :389/. If
the port number is not specified, the PT A server
attempts to connect using either of the standard
ports:
Port 389 if ldap:// is specified in the URL.
Port 636 if ldaps:// is specified in the URL.
See Section 13.6.2.2, “Specifying the
Authenticating Directory Server” for more
information.
subtree
T he pass-through subtree. T he PT A Directory
Server passes through bind requests to the
authenticating Directory Server from all clients
whose DN is in this subtree. See Section 13.6.2.3,
“Specifying the Pass-through Subtree” for more
information. T his subtree must not exist on this
server. T o pass the bind requests for
o=NetscapeRoot to the configuration directory,
the subtree o=NetscapeRoot must not exist on
the server.
maxconns
Optional. T he maximum number of connections
the PT A directory can simultaneously open to the
authenticating directory. T he default is 3. See
Section 13.6.2.4, “Configuring the Optional
Parameters” for more information.
maxops
Optional. T he maximum number of simultaneous
operations (usually bind requests) the PT A
directory can send to the authenticating directory
within a single connection. T he default is 5. See
Section 13.6.2.4, “Configuring the Optional
Parameters” for more information.
timeout
Optional. T he time limit, in seconds, that the PT A
directory waits for a response from the
authenticating Directory Server. If this timeout is
exceeded, the server returns an error to the
client. T he default is 300 seconds (five minutes).
Specify zero (0) to indicate no time limit should be
550
Chapter 13. Managing User Authentication
enforced. See Section 13.6.2.4, “Configuring the
Optional Parameters” for more information.
ldver
Optional. T he version of the LDAP protocol used
to connect to the authenticating directory.
Directory Server supports LDAP version 2 and 3.
T he default is version 3, and Red Hat strongly
recommends against using LDAPv2, which is old
and will be deprecated. See Section 13.6.2.4,
“Configuring the Optional Parameters” for more
information.
connlifetime
Optional. T he time limit, in seconds, within which a
connection may be used. If a bind request is
initiated by a client after this time has expired, the
server closes the connection and opens a new
connection to the authenticating directory. T he
server will not close the connection unless a bind
request is initiated and the directory determines
the connection lifetime has been exceeded. If this
option is not specified, or if only one host is listed,
no connection lifetime will be enforced. If two or
more hosts are listed, the default is 300 seconds
(five minutes). See Section 13.6.2.4, “Configuring
the Optional Parameters” for more information.
startT LS
Optional. A flag of whether to use Start T LS for
the connection to the authenticating directory.
Start T LS establishes a secure connection over
the standard port, so it is useful for connecting
using LDAP instead of LDAPS. T he SSL server
and CA certificates need to be available on both
of the servers.
T he default is 0, which is off. T o enable Start
T LS, set it to 1. T o use Start T LS, the LDAP URL
must use ldap:, not ldaps:.
See Section 13.6.2.4, “Configuring the Optional
Parameters” for more information.
13.6.2. Configuring the PTA Plug-in
T he only method for configuring the PT A plug-in is to modify the entry cn=Pass T hrough
Authentication, cn=plugins,cn=config. T o modify the PT A configuration:
1. Use the ldapm odify command to modify cn=Pass T hrough Authentication,
cn=plugins,cn=config.
2. Restart Directory Server.
Before configuring any of the PT A Plug-in parameters, the PT A Plug-in entry must be present in the
Directory Server. If this entry does not exist, create it with the appropriate syntax, as described in
Section 13.6.1, “PT A Plug-in Syntax”.
551
Red Hat D irectory Server 8.2 Administration Guide
NOTE
If the user and configuration directories are installed on different instances of the directory, the
PT A Plug-in entry is automatically added to the user directory's configuration and enabled.
T his section provides information about configuring the plug-in in the following sections:
Section 13.6.2.1, “Configuring the Servers to Use a Secure Connection”
Section 13.6.2.2, “Specifying the Authenticating Directory Server”
Section 13.6.2.3, “Specifying the Pass-through Subtree”
Section 13.6.2.4, “Configuring the Optional Parameters”
13.6.2.1. Configuring the Servers to Use a Secure Connection
T he PT A directory can be configured to communicate with the authenticating directory over SSL by
specifying LDAPS in the LDAP URL of the PT A directory. For example:
nsslapd-pluginarg0: ldaps://ldap.example.com:636/o=NetscapeRoot
13.6.2.2. Specifying the Authenticating Directory Server
T he authenticating directory contains the bind credentials for the entry with which the client is attempting
to bind. T he PT A directory passes the bind request to the host defines as the authenticating directory.
T o specify the authenticating Directory Server, replace authDS in the LDAP URL of the PT A directory
with the authenticating directory's hostname, as described in T able 13.4, “PT A Plug-in Parameters”.
1. Use ldapm odify edit the PT A Plug-in entry.
ldapmodify -p 389 -D "cn=Directory Manager" -w secret -h example
dn: cn=Pass Through Authentication,cn=plugins,cn=config
changetype: modify
replace: nsslapd-pluginarg0
nsslapd-pluginarg0: ldap://dirserver.example.com/o=NetscapeRoot
Optionally, include the port number. If the port number is not given, the PT A Directory Server
attempts to connect using either the standard port (389) for ldap:// or the secure port (636) for
ldaps://.
If the connection between the PT A Directory Server and the authenticating Directory Server is
broken or the connection cannot be opened, the PT A Directory Server sends the request to the
next server specified, if any. T here can be multiple authenticating Directory Servers specified, as
required, to provide failover if the first Directory Server is unavailable. All of the authentication
Directory Server are set in the nsslapd-pluginarg0 attribute.
Multiple authenticating Directory Servers are listed in a space-separate list of host:port pairs, with
this format:
ldap|ldaps://host1:port1 host2:port2/subtree
2. Restart the server.
service dirsrv restart instance_name
552
Chapter 13. Managing User Authentication
13.6.2.3. Specifying the Pass-through Subtree
T he PT A directory passes through bind requests to the authenticating directory from all clients with a
DN defined in the pass-through subtree. T he subtree is specified by replacing the subtree parameter in
the LDAP URL of the PT A directory.
T he pass-through subtree must not exist in the PT A directory. If it does, the PT A directory attempts to
resolve bind requests using its own directory contents and the binds fail.
1. Use the ldapm odify command to import the LDIF file into the directory.
ldapmodify -p 389 -D "cn=Directory Manager" -w secret -h example
dn: cn=Pass Through Authentication,cn=plugins,cn=config
changetype: modify
replace: nsslapd-pluginarg0
nsslapd-pluginarg0: ldap://dirserver.example.com/o=NetscapeRoot
For information on the variable components in this syntax, see T able 13.4, “PT A Plug-in
Parameters”.
2. Restart the server.
service dirsrv restart instance_name
13.6.2.4 . Configuring the Optional Parameters
Additional parameters the control the PT A connection can be set with the LDAP URL.
ldap|ldaps://authDS/subtree maxconns, maxops, timeout, ldver, connlifetime,
startTLS
T he maximum number of connections the PT A Directory Server can open simultaneously to the
authenticating directory, represented by maxconns in the PT A syntax. T he default value is 3.
T he maximum number of bind requests the PT A Directory Server can send simultaneously to the
authenticating Directory Server within a single connection. In the PT A syntax, this parameter is
maxops. T he default is value is 5.
T he time limit for the PT A Directory Server to wait for a response from the authenticating Directory
Server. In the PT A syntax, this parameter is timeout. T he default value is 300 seconds (five
minutes).
T he version of the LDAP protocol for the PT A Directory Server to use to connect to the authenticating
Directory Server. In the PT A syntax, this parameter is ldver. T he default is LDAPv3.
T he time limit in seconds within which a connection may be used. If a bind request is initiated by a
client after this time has expired, the server closes the connection and opens a new connection to
the authenticating Directory Server. T he server will not close the connection unless a bind request is
initiated and the server determines the timeout has been exceeded. If this option is not specified or if
only one authenticating Directory Server is listed in the authDS parameter, no time limit will be
enforced. If two or more hosts are listed, the default is 300 seconds (five minutes). In the PT A
syntax, this parameter is connlifetime.
Whether to use Start T LS for the connection. Start T LS creates a secure connection over a standard
LDAP port. For Start T LS, the servers must have their server and CA certificates installed, but they do
not need to be running in SSL.
T he default is 0, which means Start T LS is off. T o enable Start T LS, set it to 1. T o use Start T LS, the
LDAP URL must use ldap:, not ldaps:.
553
Red Hat D irectory Server 8.2 Administration Guide
1. Use ldapm odify to edit the plug-in entry.
ldapmodify -p 389 -D "cn=Directory Manager" -w secret -h example
dn: cn=Pass Through Authentication,cn=plugins,cn=config
changetype: modify
replace: nsslapd-pluginarg0
nsslapd-pluginarg0: ldap://dirserver.example.com/o=NetscapeRoot
3,5,300,3,300,0
(In this example, each of the optional parameters is set to its default value.) Make sure there is a
space between the subtree parameter, and the optional parameters.
NOTE
Although these parameters are optional, if any one of them is defined, they all must be
defined, even if they use the default values.
2. Restart the server.
service dirsrv restart instance_name
13.6.3. PTA Plug-in Syntax Examples
T his section contains the following examples of PT A Plug-in syntax in the dse.ldif file:
Section 13.6.3.1, “Specifying One Authenticating Directory Server and One Subtree”
Section 13.6.3.2, “Specifying Multiple Authenticating Directory Servers”
Section 13.6.3.3, “Specifying One Authenticating Directory Server and Multiple Subtrees”
Section 13.6.3.4, “Using Non-Default Parameter Values”
Section 13.6.3.5, “Specifying Different Optional Parameters and Subtrees for Different Authenticating
Directory Servers”
13.6.3.1. Specifying One Authenticating Directory Server and One Subtree
T his example configures the PT A Plug-in to accept all defaults for the optional variables. T his
configuration causes the PT A Directory Server to connect to the authenticating Directory Server for all
bind requests to the o=NetscapeRoot subtree. T he hostname of the authenticating Directory Server is
configdir.exam ple.com .
dn: cn=Pass Through Authentication,cn=plugins,cn=config
...
nsslapd-pluginEnabled: on
nsslapd-pluginarg0: ldap://configdir.example.com/o=NetscapeRoot
...
13.6.3.2. Specifying Multiple Authenticating Directory Servers
If the connection between the PT A Directory Server and the authenticating Directory Server is broken or
the connection cannot be opened, the PT A Directory Server sends the request to the next server
specified, if any. T here can be multiple authenticating Directory Servers specified, as required, to provide
failover if the first Directory Server is unavailable. All of the authentication Directory Server are set in the
nsslapd-pluginarg0 attribute. Multiple authenticating Directory Servers are listed in a space-separate
list of host:port pairs. For example:
554
Chapter 13. Managing User Authentication
dn: cn=Pass Through Authentication,cn=plugins,cn=config
...
nsslapd-pluginEnabled: on
nsslapd-pluginarg0: ldap://configdir.example.com:389
config2dir.example.com:1389/o=NetscapeRoot
...
NOTE
T he nsslapd-pluginarg0 attribute sets the authentication Directory Server; additional
nsslapd-pluginargN attributes can set additional suffixes for the PT A Plug-in to use, but not
additional hosts.
13.6.3.3. Specifying One Authenticating Directory Server and Multiple Subtrees
T he following example configures the PT A Directory Server to pass through bind requests for more than
one subtree (using parameter defaults):
dn: cn=Pass Through Authentication,cn=plugins,cn=config
...
nsslapd-pluginEnabled: on
nsslapd-pluginarg0: ldap://configdir.example.com/o=NetscapeRoot
nsslapd-pluginarg1: ldap://configdir.example.com/dc=example,dc=com
...
13.6.3.4 . Using Non-Default Parameter Values
T his example uses a non-default value (10) only for the maximum number of connections parameter
m axconns. Each of the other parameters is set to its default value. However, because one parameter is
specified, all parameters must be defined explicitly in the syntax.
dn: cn=Pass Through Authentication,cn=plugins,cn=config
...
nsslapd-pluginEnabled: on
nsslapd-pluginarg0: ldap://configdir.example.com/o=NetscapeRoot 10,5,300,3,300,1
...
13.6.3.5. Specifying Different Optional Parameters and Subtrees for Different Authenticating
Directory Servers
T o specify a different pass-through subtree and optional parameter values for each authenticating
Directory Server, set more than one LDAP URL/optional parameters pair. Separate the LDAP
URL/optional parameter pairs with a single space as follows.
dn: cn=Pass Through Authentication,cn=plugins,cn=config
...
nsslapd-pluginEnabled: on
nsslapd-pluginarg0:ldap://configdir.example.com/o=NetscapeRoot 10,15,30,3,600,0
nsslapd-pluginarg1:ldap://config2dir.example.com/dc=example,dc=com
7,7,300,3,300,1
...
555
Red Hat D irectory Server 8.2 Administration Guide
13.7. Using PAM for Pass-through Authentication
Pass-through authentication is when any authentication request is forwarded from one server to another
service.
Many systems already have authentication mechanisms in place for Unix and Linux users. One of the
most common authentication frameworks is Pluggable Authentication Modules (PAM). Since many
networks already existing authentication services available, administrators may want to continue using
those services. A PAM module can be configured to tell Directory Server to use an existing
authentication store for LDAP clients.
PAM pass-through authentication in Red Hat Directory Server uses the PAM Pass-through
Authentication Plug-in, which enables the Directory Server to talk to the PAM service to authenticate
LDAP clients.
Figure 13.3. PAM Pass-through Authentication Process
PAM pass-through authentication works together with account inactivation when authenticating users,
assuming that the appropriate mapping method (ENT RY) is used. However, PAM pass-through
authentication does not validate passwords against password policies set either globally or locally,
because the passwords are set and stored in the PAM module, not in the Directory Server.
13.7.1. PAM Pass-through Authentication Configuration Options
T here are three different areas that can be configured for PAM pass-through:
T he suffixes that are controlled by the PAM pass-through authentication plug-in. T his covers suffixes
to exclude, suffixes to include, and how to handle a missing suffix.
T he PAM attribute mapping. T he credentials that are offered to the Directory Server have to be
mapped in some way to an LDAP entry and then, back to the credentials in the PAM service. T his is
done by defining a mapping method and then, optionally, which LDAP attribute to use to match the
credentials.
General configuration such as using SSL connections, the PAM service to use, and whether to
fallback to LDAP authentication if PAM authentication fails.
556
Chapter 13. Managing User Authentication
T able 13.5. PAM Pass-through Auth Plug-in Attributes
Attribute
Definition
pamExcludeSuffix
Identifies suffixes to exclude from PAM
authentication.
pamIncludeSuffix
Identifies suffixes to include for PAM
authentication.
pamMissingSuffix
Identifies how to handle missing include or
exclude suffixes. T he options are ERROR (which
causes the bind operation to fail); ALLOW, which
logs an error but allows the operation to proceed;
and IGNORE, which allows the operation and
doesn't log any errors.
pamIDAttr
Sets the name of the attribute holding the PAM ID.
pamIDMapMethod
Gives the method to use to map the LDAP bind
DN to a PAM identity.
NOTE
Directory Server user account inactivation
is only validated using the ENT RY
mapping method. With RDN or DN, a
Directory Server user whose account is
inactivated can still bind to the server
successfully.
pamFallback
Sets whether to fallback to regular LDAP
authentication if PAM authentication fails.
pamSecure
Requires secure (T LS/SSL) connection for PAM
authentication.
pamService
Contains the service name to pass to PAM. T his
assumes that the service specified has a
configuration file in /etc/pam .d.
13.7.1.1. Specifying the Subtrees to T arget for PAM PT A
T he PAM PT A plug-in is applied globally, to all subtrees, by default unless they are explicitly excluded.
Excluding and including subtrees can help target what areas in the directory use PAM authentication
instead of LDAP authentication.
T he pamExcludeSuffix attribute excludes a suffix. By default, only the configuration subtree
(cn=config) is excluded. Alternatively, the PAM PT A plug-in can be applied to a subtree with the
pamIncludeSuffix attribute. Both of these attributes are multi-valued.
If the include attribute is set, for example, all other subtrees are automatically excluded. Likewise, if an
exclude attribute is set, all other suffixes are automatically included.
pamExcludeSuffix: cn=config
pamExcludeSuffix: o=NetscapeRoot
557
Red Hat D irectory Server 8.2 Administration Guide
With pamIncludeSuffix, only the given subtree is included and all others are automatically excluded.
Since this attribute is multi-valued, more than one suffix can be included in the PAM evaluation by
explicitly listing the suffixes.
pamIncludeSuffix: ou=Engineering,dc=example,dc=com
pamIncludeSuffix: ou=QE,dc=example,dc=com
T he pamMissingSuffix attribute tells the server how to handle a failure if the specified suffix (include
or exclude) doesn't exist. If it's set to IGNORE, then if the suffix doesn't exist, the plug-in simply skips
that suffix and tries the next.
pamMissingSuffix: IGNORE
pamIncludeSuffix: ou=Engineering,dc=example,dc=com
pamIncludeSuffix: ou=Not Real,dc=example,dc=com
13.7.1.2. Setting PAM PT A Mappings
T here has to be a way to connect the LDAP identity to the PAM identity. T he first thing to define is the
method to use to map the entries. T here are three options: DN, RDN, and ENT RY. ENT RY uses a userdefined attribute in the entry.
Multiple mapping methods can be supplied in an ordered, space-separated list. T he plug-in attempts to
use each mapping method in the order listed until authentication succeeds or until it reaches the end of
the list.
For example, this mapping method first maps the RDN method, then ENT RY, then DN, in the order the
methods are listed:
pamIDMapMethod: RDN ENTRY DN
T he different mapping methods are listed in T able 13.6, “Mapping Methods for PAM Authentication”.
NOTE
Directory Server user account inactivation is only validated using the ENT RY mapping method.
With RDN or DN, a Directory Server user whose account is inactivated can still bind to the server
successfully.
558
Chapter 13. Managing User Authentication
T able 13.6. Mapping Methods for PAM Authentication
Mapping
Description
RDN
T his method uses the value from the leftmost
RDN in the bind DN. T he mapping for this method
is defined by Directory Server. T his is the default
mapping method, if none is given.
ENT RY
T his method pulls the value of the PAM identity
from a user-defined attribute in the bind DN entry.
T he identity attribute is defined in the pamIDAttr
attribute.
pamIDAttr: customPamUid
DN
T his method uses the full distinguished name
from the bind DN. T he mapping for this method is
defined by Directory Server.
13.7.1.3. Configuring General PAM PT A Settings
T hree general configuration settings can be set for PAM authentication:
T he service name to send to PAM (pam Service); this is the name of the configuration file to use in
/etc/pam .d
Whether to require a secure connection (pam Secure)
Whether to fall back to LDAP authentication if PAM authentication fails (pam Fallback)
pamFallback: false
pamSecure: false
pamService: ldapserver
13.7.2. Configuring PAM Pass-through Authentication
PAM pass-through authentication is configured through the command line.
1. Make sure the PAM service is fully configured.
2. Enable the plug-in; this is disabled by default.
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389 -h
server.example.com
dn: cn=PAM Pass-through Auth Plugin,cn=plugins,cn=config
changetype: modify
replace: nsslapd-pluginEnabled
nsslapd-pluginEnabled: on
3. Add or edit the attributes available for the PAM plug-in. T he available attributes are listed in
T able 13.5, “PAM Pass-through Auth Plug-in Attributes” and Example 13.1, “Example PAM Passthrough Authentication Configuration” has an example entry.
4. Restart the server to load the plug-in.
service dirsrv restart
559
Red Hat D irectory Server 8.2 Administration Guide
Example 13.1. Example PAM Pass-through Authentication Configuration
dn: cn=PAM Pass Through Auth,cn=plugins,cn=config
objectClass: top
objectClass: nsSlapdPlugin
objectClass: extensibleObject
objectClass: pamConfig
cn: PAM Pass Through Auth
nsslapd-pluginPath: libpam-passthru-plugin
nsslapd-pluginInitfunc: pam_passthruauth_init
nsslapd-pluginType: preoperation
nsslapd-pluginEnabled: on
nsslapd-pluginloadglobal: true
nsslapd-plugin-depends-on-type: database
pamMissingSuffix: ALLOW
pamExcludeSuffix: cn=config
pamExcludeSuffix: o=NetscapeRoot
pamIDMapMethod: RDN ou=people,dc=example,dc=com
pamIDMapMethod: ENTRY ou=engineering,dc=example,dc=com
pamIDAttr: customPamUid
pamFallback: FALSE
pamSecure: TRUE
pamService: ldapserver
13.8. Inactivating Users and Roles
A single user account or set of accounts can be temporarily inactivated. Once an account is inactivated,
a user cannot bind to the directory. T he authentication operation will fail.
Users and roles are inactivated using the operational attribute nsAccountLock. When an entry contains
the nsAccountLock attribute with a value of true, the server rejects the bind.
T he same procedures are used to inactivate users and roles. However, when a role is inactivated, the
members of the role are inactivated, not the role entry itself. For more information about roles in general
and how roles interact with access control in particular, see Chapter 11, Organizing Entries with Groups,
Roles, and Views.
WARNING
T he root entry (the entry corresponding to the root or sub suffix) on a database cannot be
inactivated. Chapter 3, Creating Directory Entries has information on creating the entry for a root
or sub suffix, and Chapter 2, Configuring Directory Databases has information on creating root
and sub suffixes.
13.8.1. Activating and Inactivating Users and Roles Using the Console
All user and role entries are active by default. T hey must be manually marked inactive and, once
inactivated, must be manually re-activated.
1. Select the Directory tab.
2. Browse the navigation tree in the left navigation pane, and double-click the entry to inactivate.
560
Chapter 13. Managing User Authentication
T he Edit Entry dialog box appears.
3. Click Account in the left pane. T he right pane states that the role or user is activate. Click the
Inactivate button to inactivate the user or role (or the Activate button, to re-enable the
entry).
4. Click OK.
Alternatively, higlight the entry and select Inactivate (or Activate, if appropriate) from the Object
menu.
13.8.2. Viewing Inactive Users and Roles
1. Select the View menu, and select the Display item.
2. Select the Inactivation State item.
When the inactivation state is visible, any inactive object is listed in the right pane of the Console with a
red slash through it.
13.8.3. Inactivating and Activating Users and Roles Using the Command Line
T he Directory Server uses dual scripts to inactivate or activate entries through the command line. T he
ns-inactivate.pl and ns-activate.pl script share similar options to identify the entry to modify,
as listed in T able 13.7, “ns-inactivate.pl and ns-activate.pl Options”.
561
Red Hat D irectory Server 8.2 Administration Guide
For example, to inactivate a user account:
ns-inactivate.pl -D Directory Manager -w secret -p 389 -h example.com
"uid=jfrasier,ou=people,dc=example,dc=com"
-I
T hen, the account can be re-activated:
ns-activate.pl -D Directory Manager -w secret -p 389 -h example.com
"uid=jfrasier,ou=people,dc=example,dc=com"
-I
T able 13.7. ns-inactivate.pl and ns-activate.pl Options
Option Name
Description
-D
T he DN of the directory administrator.
-w
T he password of the directory administrator.
-p
Port used by the server.
-h
Name of the server on which the directory
resides.
-I
DN of the user account or role to inactivate or
activate, depending on the script.
For more information about running the ns-inactivate.pl and ns-activate.pl scripts, refer to
the Directory Server Configuration and Command-Line Tool Reference.
[8 ] The LDAP to o ls referenc ed in this g uid e are Mo z illa LDAP, ins talled with Direc to ry Server in the /usr/l i b 6 4 /m o zl d ap
d irec to ry o n Red Hat Enterp ris e Linux 5 (6 4-b it); d irec to ries fo r o ther p latfo rms are lis ted in Sec tio n 1.3, “ LDAP To o l Lo c atio ns ” .
Ho wever, Red Hat Enterp ris e Linux s ys tems als o inc lud e LDAP to o ls fro m O p enLDAP. It is p o s s ib le to us e the O p enLDAP
c o mmand s as s ho wn in the examp les , b ut yo u mus t us e the - x arg ument to d is ab le SASL and allo w s imp le authentic atio n.
562
Chapter 14. Configuring Secure Connections
Chapter 14. Configuring Secure Connections
By default, clients and users connect to the Red Hat Directory Server over a standard connection.
Standard connections do not use any encryption, so information is sent back and forth between the
server and client in the clear.
Directory Server supports SSL/T LS connections, Start T LS connection, and SASL authentication, which
provide layers of encryption and security that protect directory data from being read even if it is
interecepted.
14.1. Requiring Secure Connections
T he Directory Server has two different ways of encrypting data: SSL/T LS (either through an initial SSL
connection or using Start T LS to establish encryption on a previously unencrypted connection) and
SASL. If the server is configured to use it, any client can connect to the server using either encryption
method to protect data.
For additional security, the Directory Server can be configured to require a certain level of encryption
before it allows a connection. T he Directory Server can define and require a specific security strength
factor for any connection. T he SSF sets a minimum encryption level, defined by its key strength, for any
connection or operation.
T o require a minimum SSF for any and all directory operations, set the nsslapd-minssf configuration
attribute. When enforcing a minimum SSF, the Directory Server looks at each available encryption type
for an operation — SSL/T LS or SASL — and determines which has the higher SSF value and then
compares the higher value to the minimum SSF. (It is possible for both SASL authentication and SSL/T LS
to be configured for some server-to-server connections, such as replication.)
T he SSF for a connection is evaluated when the first operation is initiated on a connection. T his allows
Start T LS and SASL binds to succeed, even though those two connections initially open a regular
connection. After the T LS or SASL session is opened, then the SSF is evaluated. Any connection which
doesn't meet the SSF requirements is closed with an LDAP unwilling to perform error.
Setting a minimum SSF can be used to effectively disable all standard or unsecure connections to a
directory.
T he default nsslapd-minssf attribute value is 0, which means there is no minimum SSF for server
connections. T he value can be set to any reasonable positive integer. T he value represents the
required key strength for any secure connection.
1. Add the nsslapd-minssf attribute to the cn=config entry:
/usr/lib64/mozldap/ldapmodify -D "cn=directory manager" -w secret -p 389
dn: cn=config
changetype: modify
replace: nsslapd-minssf
nsslapd-minssf: 128
2. Restart the server.
service dirsrv restart
563
Red Hat D irectory Server 8.2 Administration Guide
TIP
An ACI can be set to require an SSF for a specific type of operation, as in Section 12.9.10,
“Setting an ACI to Require a Certain Security Strength Factor for Some Operations”.
Secure connections can be required for bind operations by turning on the nsslapd-requiresecure-binds attribute, as in Section 13.5.1, “Requiring Secure Binds”.
14.2. Using TLS/SSL
T o provide secure communications over the network, Red Hat Directory Server includes the LDAPS
communications protocol. LDAPS is the standard LDAP protocol, running over T ransport Layer Security
(T LS, formerly Secure Sockets Layer or SSL). Directory Server also allows spontaneous secure
connections over otherwise-insecure LDAP ports, using the Start T LS LDAP extended operation.
T he Directory Server supports T LS/SSL to secure communications between LDAP clients and the
Directory Server, between Directory Servers that are bound by a replication agreement, or between a
database link and a remote database. Directory Server can use T LS/SSL with simple authentication (bind
DN and password) or with certificate-based authentication.
Directory Server's cryptographic services are provided by Mozilla Network Security Services (NSS), a
library of T LS/SSL and base cryptographic functions. NSS includes a software-based cryptographic
token which is FIPS 140-2 certified.
Using T LS/SSL with simple authentication ensures confidentiality and data integrity. T here are two major
benefits to using a certificate — smart card, token, or software-based — to authenticate to the Directory
Server instead of a bind DN and password:
Improved efficiency. When using applications that prompt once for the certificate database password
and then use that certificate for all subsequent bind or authentication operations, it is more efficient
than continuously providing a bind DN and password.
Improved security. T he use of certificate-based authentication is more secure than non-certificate
bind operations because certificate-based authentication uses public-key cryptography. Bind
credentials cannot be intercepted across the network. If the certificate or device is lost, it is useless
without the PIN, so it is immune from third-party interference like phishing attacks.
T he Directory Server is capable of simultaneous T LS/SSL and non-SSL communications. T his means
that you do not have to choose between T LS/SSL or non-SSL communications for the Directory Server;
both can be used at the same time. Directory Server can also utilize the Start T LS extended operation to
allow T LS/SSL secure communication over a regular (insecure) LDAP port.
14.2.1. Enabling TLS/SSL: Summary of Steps
1. Obtain and install a certificate for the Directory Server, and configure the Directory Server to trust
the certification authority's (CA's) certificate.
For information, see Section 14.2.2, “Obtaining and Installing Server Certificates”.
2. T urn on T LS/SSL in the directory.
For information, refer to Section 14.2.3, “Configuring the Directory Server to Run in SSL/T LS”.
3. Configure the Admin Server connect to a T LS-enabled Directory Server.
4. Optionally, ensure that each user of the Directory Server obtains and installs a personal certificate
for all clients that will authenticate with T LS/SSL.
564
Chapter 14. Configuring Secure Connections
For information, refer to Section 14.2.10.1, “Configuring Directory Server to Accept CertificateBased Authentication from LDAP Clients”.
14.2.2. Obtaining and Installing Server Certificates
Before the Directory Server can be set to run in T LS/SSL, server and CA certificates must be properly
configured in the Directory Server. If a server certificate has already been generated for the Directory
Server instance and the issuing certificate authority (CA) is already trusted by the Directory Server,
begin setting up T LS/SSL as described in Section 14.2.3, “Configuring the Directory Server to Run in
SSL/T LS”.
Obtaining and installing certificates consists of the following steps:
1. Generate a certificate request.
2. Send the certificate request to a certificate authority.
3. Install the server certificate.
4. Set the Directory Server to trust the certificate authority.
5. Confirm that the certificates are installed.
T wo wizards automate the process of creating a certificate database and of installing the key-pair. T he
Certificate Request Wizard in the Directory Server Console can generate a certificate request
and send it to a certificate authority. T he Certificate Install Wizard in the Directory Server
Console can then install the server certificate and the CA certificate.
14 .2.2.1. Generating a Certificate Request
Generate a certificate request, and send it to a CA. T he Directory Server Console has a tool, the
Certificate Request Wizard, which generates a valid certificate request to submit to any
certificate authority (CA).
1. Select the T asks tab, and click Manage Certificates.
2. Select the Server Certs tab, and click the Request button.
565
Red Hat D irectory Server 8.2 Administration Guide
3. Enter the Requester Inform ation in the blank text fields, then click Next.
Server Name. Enter the fully qualified hostname of the Directory Server as it is used in DNS
and reverse DNS lookups; for example, dir.exam ple.com . T he server name is critical for
client-side validation to work, which prevents man-in-the-middle attacks.
Organization. Enter the legal name of the company or institution. Most CAs require this
information to be verified with legal documents such as a copy of a business license.
Organizational Unit. Optional. Enter a descriptive name for the organization within the
company.
Locality. Optional. Enter the company's city name.
State or Province. Enter the full name of the company's state or province (no abbreviations).
Country. Select the two-character abbreviation for the country's name (ISO format). T he
566
Chapter 14. Configuring Secure Connections
Country. Select the two-character abbreviation for the country's name (ISO format). T he
country code for the United States is US.
4. If an external security device is to be used to store the Directory Server certificates, the device is
plugged in, and the module has been installed as described in Section 14.2.8, “Using External
Security Devices”, then the module is available in the Active Encryption T oken menu. T he
default is to use the software databases with Directory Server, internal (software).
5. Enter the password that will be used to protect the private key, and click Next.
T he Next button is grayed out until a password is supplied.
6. T he Request Subm ission dialog box provides two ways to submit a request: directly to the CA
(if there is one internally) or manually. T o submit the request manually, select Copy to
Clipboard or Save to File to save the certificate request which will be submitted to the CA.
7. Click Done to dismiss the Certificate Request Wizard.
After generating the certificate request, send it to the CA.
14 .2.2.2. Sending a Certificate Request
After the certificate request is generated, send it to a certificate authority (CA); the CA will generate
return a server certificate.
1. Most certificate requests are emailed to the CA, so open a new message.
2. Copy the certificate request information from the clipboard or the saved file into the body of the
567
Red Hat D irectory Server 8.2 Administration Guide
message.
-----BEGIN NEW CERTIFICATE REQUEST----MIIBrjCCARcCAQAwbjELMAkGA1UEBhMCVXMxEzARBgNVBAgTCkNBTElGT1J
OSUExLDAqBgVBAoTI25ldHNjYXBlIGNvbW11bmljYXRpb25zIGNvcnBvcmF
0aW9uMRwwGgYDVQQDExNtZWxsb24ubmV0c2NhcGUuY29tMIGfMA0GCSqGSI
b3DQEBAQUAA4GNADCBiQKBgQCwAbskGh6SKYOgHy+UCSLnm3ok3X3u83Us7
ug0EfgSLR0f+K41eNqqRftGR83emqPLDOf0ZLTLjVGJaH4Jn4l1gG+JDf/n
/zMyahxtV7+mT8GOFFigFfuxaxMjr2j7IvELlxQ4IfZgWwqCm4qQecv3G+N
9YdbjveMVXW0v4XwIDAQABoAAwDQYK
------END NEW CERTIFICATE REQUEST-----
3. Send the email message to the CA.
After emailing the certificate request, wait for the CA to respond with the server certificate. Response
time for requests varies. For example, if the CA is internal to the company, it may only take a day or two
to respond to the request. If the selected CA is a third-party, it could take several weeks to respond to
the request.
After receiving the certificate, install it in the Directory Server's certificate database. When the CA sends
a response, be sure to save the information in a text file. T he certificate must be available to install in the
Directory Server.
Also, keep a backup of the certificate data in a safe location. If the system ever loses the certificate data,
the certificate can be reinstalled using the backup file.
14 .2.2.3. Installing the Certificate
NOTE
If an existing certificate has the same name as the new certificate you are attempting to install,
you cannot use the Directory Server Console to install the certificate. It fails with the error Internal
error: Fail to install certificate -8169.
You can use certutil to install the certificate.
certutil -importcert -v /path/to/certificate_file
For more information on using certutil to manage certificates, see Section 14.2.5, “Using
certutil”.
1. Select the T asks tab, and click Manage Certificates.
2. Select the Server Certs tab, and click Install.
568
Chapter 14. Configuring Secure Connections
3. Give the certificate location or paste the certificate text in the text box, then click Next.
In this file. Enter the absolute path to the certificate in this field.
In the following encoded text block. Copy the text from the CA's email or from the created text
file, and paste it in this field.
4. Check that the certificate information displayed is correct, and click Next.
5. Give a name to the certificate, and click Next.
6. Provide the password that protects the private key. T his password is the same as the one
provided in step 5 in Section 14.2.2.1, “Generating a Certificate Request”.
After installing the server certificate, configure the Directory Server to trust the CA which issued the
server's certificate.
569
Red Hat D irectory Server 8.2 Administration Guide
14 .2.2.4 . T rusting the Certificate Authority
Configuring the Directory Server to trust the certificate authority consists of obtaining the CA's certificate
and installing it into the server's certificate database. T his process differs depending on the certificate
authority. Some commercial CAs provide a web site that allow users to automatically download the
certificate. Others will email it back to users.
After receiving the CA certificate, use the Certificate Install Wizard to configure the Directory
Server to trust the certificate authority.
1. Select the T asks tab, and click Manage Certificates.
2. Go to the CA Certs tab, and click Install.
3. If the CA's certificate is saved to a file, enter the path in the field provided. Alternatively, copy and
paste the certificate, including the headers, into the text box. Click Next.
4. Check that the certificate information that opens is correct, and click Next.
5. Name the certificate, and click Next.
6. Select the purpose of trusting this certificate authority; it is possible to select both options.
570
Chapter 14. Configuring Secure Connections
Accepting connections from clients (Client Authentication). T he server checks that the client's
certificate has been issued by a trusted certificate authority.
Accepting connections to other servers (Server Authentication). T his server checks that the
directory to which it is making a connection (for replication updates, for example) has a
certificate that has been issued by a trusted certificate authority.
Once both the server and CA certificates are installed, it is possible to configure the Directory Server to
run in T LS/SSL. However, Red Hat recommends verifying that the certificates have been installed
correctly.
14 .2.2.5. Confirming T hat T he New Certificates Are Installed
1. Select the T asks tab, and click Manage Certificates.
2. Select the Server Certs tab.
A list of all the installed certificates for the server opens.
3. Scroll through the list. T he certificates installed previously should be listed.
It is now possible to set up the Directory Server to run in T LS/SSL.
NOTE
When renewing a certificate using the Certificate Wizard, the text on the introduction
screen does not clearly indicate that the process is renewal and not requesting a new certificate.
Also, the requester information is not filled in automatically.
14.2.3. Configuring the Directory Server to Run in SSL/TLS
Most of the time, the server should run with T LS/SSL enabled. If T LS/SSL is temporarily disabled, reenable it before processing transactions that require confidentiality, authentication, or data integrity.
Before T LS/SSL can be activated, first create a certificate database, obtain and install a server
571
Red Hat D irectory Server 8.2 Administration Guide
certificate, and trust the CA's certificate, as described in Section 14.2.2, “Obtaining and Installing Server
Certificates”.
With T LS/SSL enabled, when the server restarts, it prompts for the PIN or password to unlock the key
database. T his is the same password used when the server certificate and key were imported into the
database. Restarting the Directory Server without the password prompt is possible by using use a
hardware crypto device or creating a PIN file (Section 14.2.3.3, “Creating a Password File for the
Directory Server”).
WARNING
T he Directory Server must already be configured to run in SSL (Section 14.2.3.1, “Enabling
T LS/SSL Only in the Directory Server”) and the server must already have been restarted
before the Directory Server Console can be configured to use SSL. Configuring SSL/T LS for the
server requires a server restart to load the new configuration, including the new secure port
assignment. However, enabling SSL/T LS for the Console takes effect immediately. T herefore, if
the Console has SSL/T LS enabled before the server is running in SSL/T LS, then the Console
loses the connection to the server and cannot reconnect.
T o disable SSL/T LS in the Directory Server Console, use ldapm odify to edit the
nsServerSecurity attribute:
nsServerSecurity: off
NOTE
On T LS-enabled servers, be sure to check the file permissions on certificate database files, key
database files, and PIN files to protect the sensitive information they contain. Because the server
does not enforce read-only permissions on these files, check the file modes to protect the
sensitive information contained in these files.
T he files must be owned by the Directory Server user, such as the default nobody. T he key and
cert databases should be owned by the Directory Server user and should typically have
read/write access for the owner with no access allowed to any other user (mode 0600). T he PIN
file should also be owned by the Directory Server user and set to read-only by this user, with no
access to anyone other user (mode 04 00).
14 .2.3.1. Enabling T LS/SSL Only in the Directory Server
1. Obtain and install CA and server certificates.
2. Set the secure port for the server to use for T LS/SSL communications. In the Configuration
area, select the Settings tab, and enter the value in the Encrypted Port field.
572
Chapter 14. Configuring Secure Connections
T he encrypted port number must not be the same port number used for normal LDAP
communications. By default, the standard port number is 389, and the secure port is 636.
3. Select the Configuration tab, and then select the top entry in the navigation tree in the left
pane. Select the Encryption tab in the right pane.
4. Select the Enable SSL for this Server checkbox.
5. Check the Use this Cipher Fam ily checkbox.
6. Select the certificate to use from the drop-down menu.
7. Click Cipher Settings.
573
Red Hat D irectory Server 8.2 Administration Guide
By default, all ciphers are selected.
8. Set the preferences for client authentication.
Do not allow client authentication. With this option, the server ignores the client's certificate.
T his does not mean that the bind will fail.
Allow client authentication. T his is the default setting. With this option, authentication is
performed on the client's request. For more information about certificate-based authentication,
see Section 14.2.10, “Using Certificate-Based Authentication”.
Require client authentication. With this option, the server requests authentication from the
client.
If T LS/SSL is only enabled in the Directory Server and not the Directory Server Console, do not
select Require client authentication checkbox.
NOTE
T o use certificate-based authentication with replication, the consumer server must be
configured either to allow or to require client authentication.
9. T o verify the authenticity of requests, select the Check hostnam e against nam e in
certificate for outbound SSL connections option. T he server does this verification by
matching the hostname against the value assigned to the common name (cn) attribute of the
subject name in the being presented for authentication.
574
Chapter 14. Configuring Secure Connections
By default, this feature is disabled. If it's enabled and if the hostname does not match the cn
attribute of the certificate, appropriate error and audit messages are logged. For example, in a
replicated environment, messages similar to these are logged in the supplier server's log files if it
finds that the peer server's hostname doesn't match the name specified in its certificate:
[DATE] - SSL alert: ldap_sasl_bind("",LDAP_SASL_EXTERNAL) 81 (Netscape
runtime error -12276 Unable to communicate securely with peer: requested domain name does not
match the server's
certificate.)
[DATE] NSMMReplicationPlugin - agmt="cn=to ultra60 client auth"
(ultra60:1924): Replication
bind with SSL client authentication failed: LDAP error 81 (Can't contact
LDAP server)
Red Hat recommends enabling this option to protect Directory Server's outbound SSL connections
against a man-in-the-middle (MIT M) attack.
10. Click Save.
11. Restart the Directory Server. T he Directory Server must be restarted from the command line.
service dirsrv restart instance
When the server restarts, it prompts for the PIN or password to unlock the key database. T his is
the same password used when the server certificate and key were imported into the database.
T o restart the Directory Server without the password prompt, create a PIN file or use a hardware
crypto device. See Section 14.2.3.3, “Creating a Password File for the Directory Server” for
information on how to create a PIN file.
14 .2.3.2. Enabling T LS/SSL in the Directory Server, Admin Server, and Console
1. Obtain server certificates and CA certs, and install them on the Directory Server. T his is described
in Section 14.2.2, “Obtaining and Installing Server Certificates”.
2. Obtain and install server and CA certificates on the Admin Server. T his is a similar process as for
the Directory Server.
NOTE
It is important that the Admin Server and Directory Server have a CA certificate in common
so that they can trust the other's certificates.
3. Configure T LS/SSL for the Directory Server as described in Section 14.2.3.1, “Enabling T LS/SSL
Only in the Directory Server”.
4. Require SSL/T LS to connect to the Directory Server Console.
575
Red Hat D irectory Server 8.2 Administration Guide
WARNING
T he Directory Server must already be configured to run in SSL and the server must
already have been restarted before the Directory Server Console can be configured to
use SSL. Configuring SSL/T LS for the server requires a server restart to load the new
configuration, including the new secure port assignment. However, enabling SSL/T LS for
the Console takes effect immediately. T herefore, if the Console has SSL/T LS enabled
before the server is running in SSL/T LS, then the Console loses the connection to the
server and cannot reconnect.
T o disable SSL/T LS in the Directory Server Console, use ldapm odify to edit the
nsServerSecurity attribute:
nsServerSecurity: off
a. Reopen the Directory Server Console.
b. In the Configuration tab, select the server, and open the Encryption tab.
c. Check the Use SSL in the Console box.
5. In the Admin Server Console, select the Configuration tab. Select the Encryption tab, check
the Enable SSL checkbox, and fill in the appropriate certificate information.
6. In the Configuration DS tab, change the port number to the new Directory Server secure port
information. See Section 1.7, “Changing Directory Server Port Numbers” for more information. Do
this even if the default port of 636 is used. Check the Secure Connection checkbox.
576
Chapter 14. Configuring Secure Connections
7. In the User DS tab, select the Set User Directory radio button, and fill in the Directory
Server secure port information, the LDAP URL, and the user database information. Check the
Secure Connection checkbox.
8. Save the new T LS/SSL settings and Configuration DS and User DS information in the Admin
Server Console.
9. Restart the Admin Server. T he server must be restarted from the command line.
service dirsrv-admin restart
When the server restarts, it prompts for the PIN or password to unlock the key database. T his is
the same password used when the server certificate and key were imported into the database.
T o restart the Admin Server without the password prompt, create a PIN file or use a hardware
crypto device. See Section 14.2.3.4, “Creating a Password File for the Admin Server” for
information on how to create a PIN file.
577
Red Hat D irectory Server 8.2 Administration Guide
NOTE
When next logging into the Directory Server Console, be certain that the address reads https;
otherwise, the operation will time out, unable to find the server since it is running on a secure
connection. After successfully connecting, a dialog box appears to accept the certificate. Click OK
to accept the certificate (either only for that current session or permanently).
14 .2.3.3. Creating a Password File for the Directory Server
It is possible to store the certificate password in a password file. By placing the certificate database
password in a file, the server can be started from the Directory Server Console and also restarted
automatically when running unattended.
WARNING
T his password is stored in clear text within the password file, so its usage represents a
significant security risk. Do not use a password file if the server is running in an unsecured
environment.
T he password file must be in the same directory where the other key and certificate databases for
Directory Server are stored. T his is usually the main configuration directory,
/etc/dirsrv/slapd-instance_name. T he file should be named pin.txt.
Include the token name and password in the file. For example:
Internal (Software) Token:secret
For the NSS software crypto module (the default software database), the token is always called
Internal (Software) T oken.
T he PIN file should be owned by the Directory Server user and set to read-only by the Directory Server
user, with no access to anyone other user (mode 04 00).
14 .2.3.4 . Creating a Password File for the Admin Server
Like the Directory Server, the Admin Server can use a password file during login when T LS/SSL is
enabled.
578
Chapter 14. Configuring Secure Connections
WARNING
T his password is stored in clear text within the password file, so its usage represents a
significant security risk. Do not use a password file if the server is running in an unsecured
environment.
1. Open the Admin Server configuration directory, /etc/dirsrv/adm in-serv.
2. Create a password file named password.conf. T he file should include a line with the token
name and password, in the form token:password. For example:
internal:secret
For the NSS software crypto module (the default software database), the token is always called
internal.
T he password file should be owned by the Admin Server user and set to read-only by the Admin
Server user, with no access to any other user (mode 04 00).
NOTE
T o find out what the Admin Server user ID is, run grep in the Admin Server configuration
directory:
cd /etc/dirsrv/admin-serv
grep \^User console.conf
3. In the /etc/dirsrv/adm in-serv directory, edit the nss.conf file to point to the location of the
new password file.
#
Pass Phrase Dialog:
#
Configure the pass phrase gathering process.
#
The filtering dialog program (`builtin' is a internal
#
terminal dialog) has to provide the pass phrase on stdout.
NSSPassPhraseDialog file://etc/dirsrv/admin-serv/password.conf
4. Restart the Admin Server.
service dirsrv-admin restart
14.2.4. Command-Line Functions for Start TLS
LDAP operations such as ldapm odify, ldapsearch, and ldapdelete can use T LS/SSL when
communicating with an SSL-enabled server or to use certificate authentication. Command-line options
also specify or enforce Start T LS, which which allows a secure connection to be enabled on a clear text
port after a session has been initiated.
579
Red Hat D irectory Server 8.2 Administration Guide
IMPORTANT
T hese options to use Start T LS applies only for the Mozilla LDAP tools provided with Red Hat
Directory Server.
In the following example, a network administrator enforces Start T LS for a search for Mike Connor's
identification number:
ldapsearch -p 389 -ZZZ -P certificateDB -s base
-b "uid=mconnors,ou=people,dc=example,dc=com" "(attribute=govIdNumber)"
-ZZZ enforces Start T LS, and certificateDB gives the filename and path to the certificate database.
NOTE
T he -ZZZ option enforces the use of Start T LS, and the server must respond that a Start T LS
command was successful. If the -ZZZ command is used and the server does not support Start
T LS, the operation is aborted immediately.
For information on the command-line options available, see the Directory Server Configuration and
Command-Line Tool Reference.
With the -ZZ option, the following errors could occur:
If there is no certificate database, the operation fails. See Section 14.2.2, “Obtaining and Installing
Server Certificates” for information on using certificates.
If the server does not support Start T LS, the connection proceeds in clear text. T o enforce the use of
Start T LS, use the -ZZZ command option.
If the certificate database does not have the certificate authority (CA) certificate, the connection
proceeds in clear text. See Section 14.2.2, “Obtaining and Installing Server Certificates” for
information on using certificates.
With the -ZZZ option, the following errors could occur, causing the Start T LS operation to fail:
If there is no certificate database. See Section 14.2.2, “Obtaining and Installing Server Certificates”
for information on using certificates.
If the certificate database does not have the certificate authority (CA) certificate. See Section 14.2.2,
“Obtaining and Installing Server Certificates” for information on using certificates.
T he server does not support Start T LS as an extended operation.
For SDK libraries used in client programs, if a session is already in T LS mode and Start T LS is
requested, then the connection continues to be in secure mode but prints the error "DSA is
unwilling to perform ".
14.2.5. Using certutil
T he Directory Server has a command-line tool, certutil, which locally creates self-signed CA and
client certificates, certificate databases, and keys. T he default location for the Directory Server certutil
tool is /usr/bin.
580
Chapter 14. Configuring Secure Connections
certutil can also be downloaded from ftp://ftp.mozilla.org/pub/mozilla.org/security/nss/releases/.
14 .2.5.1. Creating Directory Server Certificates through the Command Line
T he following steps outline how to make the databases, key, CA certificate, server/client certificate, and
convert the certificates into pkcs12 format.
1. Open the directory where the Directory Server certificate databases are stored.
cd /etc/dirsrv/slapd-instance_name
2. Make a backup copy of all of the filed in the directory as a precaution. If something goes awry with
while managing certificates, the databases can then be restored. For example:
tar -cf /tmp/db-backup.tar *
3. Create a password file for the security token password.
vi /tmp/pwdfile
secretpw
T his password locks the server's private key in the key database and is used when the keys and
certificates are first created. T he password in this file is also the default password to encrypt
PK12 files used by pk12util. Because this password is stored in plaintext, the password file
should be owned by the user as which Directory Server runs, by default nobody, and it must be
set as read-only for the Directory Server user and allow no access to anyone else (mode 04 00).
It's a good idea to have a secure backup of this file.
4. Set the environment variable for the shell to include the certutil directory path. For example:
export PATH=/usr/bin/:$PATH
T he command varies depending on the shell.
5. Create the key and certificate databases databases.
certutil -N -d . -f /tmp/pwdfile
6. Generate the self-signed CA certificate. certutil creates the required key pairs and the
certificate. T his certificate is used to generate the other server certificates and can be exported
for use with other servers and clients.
certutil -S -n "CA certificate" -s "cn=My Org CA cert,dc=example,dc=com" -2
-x -t "CT,," -m 1000 -v 120 -d . -k rsa -f /tmp/pwdfile
7. Generate the Directory Server client certificate.
certutil -S -n "Server-Cert" -s "cn=FQDN" -c "CA certificate" -t "u,u,u" -m
1001 -v 120 -d . -k rsa -f /tmp/pwdfile
T he value of the -s argument is very important. T he leftmost RDN must be cn=FQDN (where
FQDN is the fully-qualified host and domain name of the Directory Server). For example, to issue a
certificate for a server with the name ldap.exam ple.com , specify at least -s
"cn=ldap.exam ple.com "; it is beneficial to have a more descriptive name to help with server
identification, such as "cn=ldap.exam ple.com ,ou=DS1". T he FQDN must be available for
581
Red Hat D irectory Server 8.2 Administration Guide
DNS and reverse DNS lookups to Directory Server clients because certificate validation may fail if
the clients cannot properly resolve the FQDN, and some clients refuse to connect if a server
certificate does not have its FQDN in the subject. Additionally, using the format
cn=hostname.domain is essential for Directory Server clients to protect themselves from man in
the middle attacks.
NOTE
T here should only be one cn in a certificate subject name. T o add detail to the subject
name, use cn as the RDN and another attribute — like ou, l, or c — for the other subject
name elements.
T o provide a subjectAltName, as well as the nickname, use the -8 argument in addition to the -s
argument.
T o use the Directory Server behind a DNS round robin or any other scheme which aliases a
single server certificate to multiple hostnames, see the T LS/SSL information about server name
wildcards or subjectAltName.
Server certificates for other servers are created using a similar command as for the Directory
Server certificate. Make sure that every -n option (nickname) and -m option (serial number) is
unique for every certificate, and make sure that the -s option gives the correct FQDN for the
server.
NOTE
Keep careful track on the numbers set with the -m option. T he -m option sets the unique
identifier for the server certificate, and a CA cannot issue two certificates with the same ID.
Keep a log of issued serial numbers so that no number is ever duplicated.
8. Export the CA certificate for use with other servers and clients. A client usually requires the CA
certificate to validate the server certificate in an T LS/SSL connection. Use certutil to export the
CA certificate in ASCII/PEM format:
certutil -d . -L -n "CA certificate" -a > cacert.asc
T he way that the CA certificate is imported is different for every client. For example, certutil
can import a CA certificate into another Directory Server certificate database:
cd /etc/dirsrv/slapd-otherserver
certutil -A -d . -n "CA certificate" -t "CT,," -a -i cacert.asc
9. Use pk12util to export other server certificates and keys created with certutil so that they
can be used on a remote server.
pk12util -d . -o ldap1.p12 -n Server-Cert -w /tmp/pwdfile -k /tmp/pwdfile
T he -w argument is the password used to encrypt the .p12 file for transport. T he -k argument
specifies the password for the key database containing the server certificate being exported to
.p12.
10. If the Directory Server will run with T LS/SSL enabled, then create a password file (pin.txt) for
the server to use so it will not prompt you for a password every time it restarts. Creating the
582
Chapter 14. Configuring Secure Connections
password file is described in Section 14.2.3.3, “Creating a Password File for the Directory Server”.
T he certificates created by certutil are automatically available in the Encryption tab of the
Console. T here is no need to import them because they are already in the certificate database.
14 .2.5.2. certutil Usage
certutil can be used for a variety of tasks to manage certificates and keys, such as generating
certificate requests and removing certificates from the certificate database. Some of the most common
options are listed in T able 14.1, “certutil Options”. For the full list of commands and arguments, run
certutil -H from the command line.
583
Red Hat D irectory Server 8.2 Administration Guide
T able 14 .1. certutil Options
Options
Description
-x
Creates a self-signed CA certificate.
-S
Creates a server or client certificate.
-R
Generates a certificate request.
-N
Creates new security databases.
-L
Lists all of the certificates in the database.
-A
Adds a certificate to the certificate database.
-n
Gives the name of the certificate.
-d
Certificate database directory; this is the directory
for the subsystem instance.
-m
T he serial number for the certificate.
-k
T he key type to use; the only option is rsa.
-g
T he key size. T he recommended size for RSA
keys is 2048.
-s
T he subject name of the certificate.
-t
T he trust arguments for the certificate, meaning
the purposes for which the certificate is allowed
to be used.
-v
T he validity period, in months.
numbers 1-8
T hese set the available certificate extensions.
Only eight can be specified through the
certutil tool:
Key Usage: 1
Basic Constraints: 2
Certificate Authority Key ID: 3
CRL Distribution Point: 4
Netscape Certificate T ype: 5
Extended Key Usage: 6
Email Subject Alternative Name: 7
DNS Subject Alternative Name: 8
-a
Outputs the certificate request to an ASCII file
instead of binary.
-o
T he output file to which to save the certificate
request.
-i
An input file containing a certificate.
-f
T he path to a password file for the security
databases password.
T able 14.2, “certutil Examples” has some of the common uses for the certutil command.
584
Chapter 14. Configuring Secure Connections
T able 14 .2. certutil Examples
Example
Description
certutil -L -d .
Lists the certificates in the database.
certutil -N -d .
Creates new key (key3.db) and certificate
(cert8.db) databases.
certutil -S -n "CA certificate" -s "cn=My Org CA
cert,dc=example,dc=com" -2 -x -t "CT ,," -m 1000
-v 120 -d . -k rsa
Creates a self-signed CA certificate.
certutil -S -n "Server-Cert" -s "cn=FQDN" -c "CA
certificate" -t "u,u,u" -m 1001 -v 120 -d . -k rsa
Creates a client certificate.
certutil -L -d . -n "cert_name"
"Pretty prints" the specified certificate; the
cert_name can specify either a CA certificate or a
client certificate.
certutil -L -d . -n "cert_name" > certfile.asc
Exports the specified certificate out of the
database to ASCII (PEM) format.
certutil -L -d . -n "cert_name" -r > certfile.bin
Exports the specified certificate out of the
database to binary format; this can be used with
Directory Server attributes such as
usercertificate;binary.
14.2.6. Managing Certificates Used by the Directory Server Console
T he certificates and keys used by the server are stored in NSS security databases in the
/etc/dirsrv/slapd-instance_name directory. T he Directory Server Console itself also uses
certificates and keys for SSL/T LS connections; these certificates are stored in a separate database in a
separate location, $HOME/.redhat-idm -console. If the Directory Server Console is used to connect
to multiple instances of Directory Server over SSL, then it is necessary to trust every CA which issued
the certificates for every Directory Server instance.
When SS">