SAS 9.4 Security Administration Guide Intelligence Platform User Guide
Reklaam
Reklaam
SAS
®
9.4 Intelligence
Platform: Security
Administration Guide, Third
Edition
SAS
®
Documentation
The correct bibliographic citation for this manual is as follows: SAS Institute Inc. 2016. SAS® 9.4 Intelligence Platform: Security Administration
Guide, Third Edition. Cary, NC: SAS Institute Inc.
SAS® 9.4 Intelligence Platform: Security Administration Guide, Third Edition
Copyright © 2016, SAS Institute Inc., Cary, NC, USA
All Rights Reserved. Produced in the United States of America.
For a hard copy book: No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, or otherwise, without the prior written permission of the publisher, SAS Institute Inc.
For a web download or e-book: Your use of this publication shall be governed by the terms established by the vendor at the time you acquire this publication.
The scanning, uploading, and distribution of this book via the Internet or any other means without the permission of the publisher is illegal and punishable by law. Please purchase only authorized electronic editions and do not participate in or encourage electronic piracy of copyrighted materials. Your support of others' rights is appreciated.
U.S. Government License Rights; Restricted Rights: The Software and its documentation is commercial computer software developed at private expense and is provided with RESTRICTED RIGHTS to the United States Government. Use, duplication, or disclosure of the Software by the
United States Government is subject to the license terms of this Agreement pursuant to, as applicable, FAR 12.212, DFAR 227.7202-1(a), DFAR
227.7202-3(a), and DFAR 227.7202-4, and, to the extent required under U.S. federal law, the minimum restricted rights as set out in FAR
52.227-19 (DEC 2007). If FAR 52.227-19 is applicable, this provision serves as notice under clause (c) thereof and no other notice is required to be affixed to the Software or documentation. The Government’s rights in Software and documentation shall be only those set forth in this Agreement.
SAS Institute Inc., SAS Campus Drive, Cary, NC 27513-2414
November 2016
SAS® and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of SAS Institute Inc. in the USA and other countries. ® indicates USA registration.
Other brand and product names are trademarks of their respective companies.
9.4-P1:bisecag
Contents
About This Document . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
What's New in Security Administration in SAS 9.4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Accessibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Fundamentals
Configuring Client Machines to Use Integrated Windows Authentication . . . . . . . . . . 23
Authorization
iv
Contents
Variation 1: Regional Separation, Designated Content Creators . . . . . . . . . . . . . . . . . . 72
Authentication
Contents
v
Encryption
How to Change Encoding for Data in Motion for IOM Servers . . . . . . . . . . . . . . . . . 202
How to Change Encoding for Data in Motion for Outbound Passwords . . . . . . . . . . . 205
How to Configure TLS between the Metadata Server and an LDAP Server . . . . . . . . 207
How to Enable IOM Communication over HTTPS Proxy . . . . . . . . . . . . . . . . . . . . . . 219
Appendix
vi
Contents
Location of the User Bulk Load and Synchronization Macros . . . . . . . . . . . . . . . . . . 244
Recommended Reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
vii
About This Document
Audience
This document helps SAS administrators understand the suite-wide security features of the platform.
This document is organized as follows:
• The first part, Fundamentals, contains the essential information for all security administrators:
• a brief overview of security features
• an orientation to managing SAS metadata identities (users, groups, and roles)
• an orientation to the SAS metadata authorization layer (permissions)
• instructions for selected tasks
• The rest of the document contains reference information and instructions for specialized configurations.
Here is a list of related topics that are outside the scope of this book:
• Security features that are provided by the middle tier are documented in
SAS
Intelligence Platform: Middle-Tier Administration Guide
.
• Security features that are specific to a particular product are documented in the administrative guide for that product. For example, roles and capabilities are documented as application-specific features.
• Comprehensive information about each implementation of fine-grained access to data is included in documentation for the implementation’s underlying technology.
See “Fine-Grained Controls for Data” on page 62 .
• For assistance in using a particular interface to perform security-related tasks, see the documentation for that interface.
Requirements
This document assumes the following conditions:
• You are familiar with the operating systems in your environment.
• You have completed all installation, deployment, and (if applicable) migration tasks.
viii
About This Document
• You are familiar with the concepts and terminology that are introduced in
SAS
Intelligence Platform: Overview
.
ix
What's New in Security
Administration in SAS 9.4
Overview
New and enhanced features in the following areas increase security and manageability:
• access to resources
• authentication
• encryption
• user administration
• documentation enhancements
Access to Resources
• You can use either SAS Management Console or the AUTHLIB procedure to bind
SAS data to metadata. All access from SAS to metadata-bound data is subject to
metadata-layer permissions. See “Metadata-Bound Libraries” on page 94
.
• You can limit the reach of a SAS server process, ensuring that the process can access only designated resources in the host operating environment. This functionality is
available in the first maintenance release for SAS 9.4. See “Locked-Down Servers” on page 96 .
• You can manage metadata-layer permissions using batch tools. The tools support viewing and setting access controls and viewing, creating, and modifying access control templates. See
Chapter 9, “Batch Tools for Metadata Authorization,” on page
Authentication
• In direct LDAP authentication, the following enhancements are available:
• You can authenticate against multiple LDAP servers.
• You can authenticate against multiple LDAP groups within an LDAP server.
x
Security Administration
• You can use the AD_STYLE environment variable to limit authentication activity to one attempt per logon interaction. By specifying which user ID format the metadata server uses in an authentication request against LDAP, you avoid inadvertent locking of Active Directory accounts due to repeated attempts using different formats for the submitted user ID.
• You can use the AD_TIMEOUT or LDAP_TIMEOUT environment variable to define how long the first attempt to connect has before it times out.
See “How to Configure Direct LDAP Authentication” on page 160 .
• In SAS internal authentication, any server-level internal account policies that have been explicitly set in the metadata server’s omaconfig.xml file are displayed in SAS
Management Console. The display is Read-Only. See “How to Change Internal
Account Policies” on page 177 .
• In the middle tier, the remote services component is no longer used. This simplifies administration and increases availability. See
“Web Authentication” on page 149
.
• In Integrated Windows authentication, only the standard service principal name format (SAS/machine) is automatically registered. A deprecated alternate format that includes a port number is no longer automatically registered. See
Integrated Windows Authentication” on page 165
.
• In the second maintenance release for SAS 9.4, Integrated Windows Authentication on Linux systems no longer requires the use of Quest Authentication Services. SAS can leverage the libraries that are shipped with the supported operating system or that are provided in most third-party authentication solutions.
• The second maintenance release for SAS 9.4 supports outbound and trusted authentication domains. A login in an outbound domain can establish connections only to other servers. For outbound domain logins, the constraints on the user ID are not enforced. A login in a trusted domain can be accessed only by a trusted system account on behalf of the user. A login in a trusted domain also requires the
authentication domain to be an outbound domain. See “Outbound and Trusted
Authentication Domains” on page 124 .
• Beginning in the third maintenance release for SAS 9.4, additional security is provided for SAS web applications. In scenarios where applications are using the
SAS middle tier as a proxy for accessing external URLs, additional security has been added through a whitelist, or security filter, and logging. You can also whitelist certain HTTP request methods.
SAS web applications that require external access to URLs must also have a whitelist of URLs that can be accessed.
For additional information, see:
• “Whitelist of Websites and Methods Allowed to Link to SAS Web Applications” in SAS Intelligence Platform: Middle-Tier Administration Guide
• “Configuring the Cross Domain Proxy Servlet through a Whitelist” in SAS
Intelligence Platform: Middle-Tier Administration Guide
• “Enabling Support for Forward Proxy Authentication” in SAS Intelligence
Platform: Middle-Tier Administration Guide
Encryption
xi
Encryption
• Starting at SAS 9.4, the SAS/SECURE product is included with Base SAS software.
In prior releases, SAS/SECURE was an add-on product that was licensed separately.
This change makes strong encryption available in all deployments (except where prohibited by import restrictions). To learn about SAS/SECURE, see
Encryption in
SAS
.
• You can use an industry standard algorithm (AES) to encrypt SAS data on disk. See
“ENCRYPT= Data Set Option” in SAS Data Set Options: Reference .
• You can use a new encoding type, SAS004 (AES encryption with 64-bit salt). The salt value of SAS004 is higher than that of SAS003 (AES with 16-bit encryption), so
using SAS004 increases security for stored passwords. See Chapter 13, “Encryption
• You can use the SAS Deployment Manager or the METADATA procedure to reencrypt passwords that are stored in metadata. This feature provides an easy way to convert currently stored passwords from one encryption type to another. This feature
is also used to introduce, change, or remove a passphrase key. See “How to Re-
Encrypt Stored Passwords” on page 199
.
• You can specify a passphrase to use as the encryption key for passwords that are stored in metadata. You supply and control the passphrase, so this feature increases
security, compared to using the default fixed key. See “How to Re-Encrypt Stored
.
• In Transport Layer Security (TLS) between the metadata server and LDAP, the
NOSSLNAMECHECK environment variable disables server name checking. This can be useful in accommodating connections to an LDAP load balancer. See
“How to Configure TLS between the Metadata Server and an LDAP Server” on page 207 .
• In the third maintenance release for SAS 9.4, the following changes have been made to certificates, Transport Layer Security (TLS), and HTTPS security:
• SAS provides a default truststore (the jssecacerts file) that takes precedence over the previous truststore (the cacerts file) in the SAS Private JRE.
The trusted CA bundle is a copy of the Mozilla bundle, which is the list of certificate authority (CA) certificates that are distributed with Mozilla software products. For more information, see https://wiki.mozilla.org/CA:IncludedCAs .
• On UNIX, trusted certificates are now located in the trusted CA bundle in
SAS-
installation-directory/
SASSecurityCertificateFramework/1.1/cacerts/ trustedcerts.pem
. During installation, SAS Deployment Wizard sets the
SSLCALISTLOC system option in the
SAS-installation-directory/
SASFoundation/9.4/sasv9.cfg
file to point to the trustedcerts.pem file by default.
• If certificates are not used, SAS ignores SSLCALISTLOC.
For more information about the order that SAS uses when searching for certificates, see Details .
• Certificate bundle management has been added to SAS Deployment Manager.
xii
Security Administration
Using SAS Deployment Manager, you can add and remove certificates to and from the trusted CA bundle. For more information, see “Add Your Certificates to the Trusted CA Bundle” in Encryption in SAS .
To learn about certificates, see “What Are Certificates?” on page 186 .
• In the fourth maintenance release of SAS 9.4, the OpenSSL libraries provided by
SAS have been updated. For SAS 9.4 and all maintenance releases of SAS 9.4, updated versions of OpenSSL for UNIX and z/OS are provided and updated through hot fixes. See the SAS Security Bulletin on OpenSSL for the most current information about the versions of OpenSSL used in SAS products and about the advisories under consideration.
For a quick reference of the OpenSSL version supported for each version of SAS
Foundation, see Mapping Between SAS Version and OpenSSL Version .
Note: On Windows deployments, SAS uses the code delivered with Windows for
TLS. On UNIX and z/OS deployments, SAS redistributes a copy of OpenSSL.
User Administration
• In SAS Management Console, you can no longer change the name of an existing user, group, or role. You can still change the display name.
• In the SAS metadata, you can no longer give an identity two logins in the same authentication domain by entering the same user ID in two different formats. See
“Uniqueness Requirements” on page 22 .
• Restricted user administrators can no longer add or update logins and internal accounts for unrestricted users.
Documentation Enhancements
In the fourth maintenance release of SAS 9.4, information has been moved from the SAS
9.4 Intelligence Platform: Installation and Configuration Guide, Second Edition to this
document. The new sections are as follows:
•
“What Are Certificates?” on page 186
•
“How SAS Uses Certificates” on page 188
•
“Obtaining Certificates” on page 192
•
“Implementing Certificates” on page 195
•
“How to Implement Certificates” on page 210
Accessibility
For information about the accessibility of any of the products mentioned in this document, see the usage documentation for that product.
xiii
xiv
Security Administration
Part 1
Fundamentals
Chapter 1
Security Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Chapter 2
User Administration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Chapter 3
Access Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Chapter 4
Selected Tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1
2
Chapter 1
Security Overview
Security in the Platform
The platform provides security features including auditing, authentication, authorization, encryption, and role-based access to application features.
The platform cooperates with systems such as the host environment, the web realm, and third-party databases.
To coordinate identity information, at least one user ID (such as a host, Active Directory, or LDAP user ID) is stored in the SAS metadata for each registered user.
Auditing
Security-related events are logged as part of a system-wide logging facility. The following table describes security-related log categories.
Table 1.1 Selected Security-Related Log Categories
Category
Audit.Authentication
Events Captured
Authentication events, client connection information.
3
4
Chapter 1 • Security Overview
Category
Audit.Meta.Security.AccCtrlAdm
Audit.Meta.Security.GrpAdm
Audit.Meta.Security.UserAdm
Events Captured
Changes to explicit controls, ACTs, application of ACTs to objects, and passwords (on objects such as Tables,
Connections, and ProtectedPassthrus). Includes additions, deletions, modifications, and failed attempts to perform these actions.
Changes to memberships (for groups or roles). Includes adding members, removing members, and failed attempts to perform these actions.
Changes to users, groups, roles, logins, internal accounts, and authentication domains. Includes additions, deletions, modifications, and failed attempts to perform these actions.
Audit.Meta.Updates.PublicObjects
Add, update, delete, and change management events on public objects.
For more information, see “Administering Logging for SAS Servers” in SAS Intelligence
Platform: System Administration Guide
.
T I P
For a collection of utilities to support reporting against the generated logs, search for "Audit and Performance Measurement" at support.sas.com
.
See Also
•
SAS Logging: Configuration and Programming Reference
• “Administering Logging for SAS Servers” in SAS Intelligence Platform: System
Administration Guide
Authorization
The platform provides a metadata-based authorization layer that supplements protections from the host environment and other systems. You can use this layer to manage access to almost any metadata object (for example, reports, data definitions, information maps, jobs, stored processes, and server definitions).
Across authorization layers, protections are cumulative. In order to perform a task, a user must have sufficient access in all applicable layers.
In the metadata layer, the following permissions are always enforced:
• the ReadMetadata permission (RM), which controls the ability to see an object
• the WriteMetadata permission (WM), which controls the ability to update or delete an object
Other permissions are specialized and affect only certain types of objects.
CAUTION:
Roles
5
In the metadata authorization layer, not all permissions are enforced for all items.
It is essential to understand which actions are controlled by each permission.
See “Use and Enforcement of Each Permission” on page 67 .
CAUTION:
Some clients enable power users to create and run SAS programs that access data directly, bypassing metadata-layer controls.
It is essential to manage
See Also
Appendix 3, “Security Report Macros,” on page 253
Encryption
The platform offers encryption features that help protect information about disk and in transit. Here is an overview of encryption support:
• Passwords in configuration files and the metadata are encrypted or encoded. Most other metadata is not encrypted.
• Passwords in transit to and from SAS servers are encrypted or encoded. You can choose to encrypt all such traffic, instead of encrypting only credentials.
• When you obtain and implement certificates for SAS Web Server and other middletier components, you can use auto-generated certificates from SAS Deployment
Wizard or provide your own.
See Also
•
Chapter 13, “Encryption Model,” on page 183
•
Chapter 14, “Encryption Tasks,” on page 197
•
Encryption in SAS
Roles
The platform supports metadata-based application roles that control which application features (such as buttons, tabs, and menu items) are visible to which users.
For example, role memberships determine who can see the Server Manager plug-in (in
SAS Management Console), compare data (in SAS Enterprise Guide), or directly open an information map (in SAS Web Report Studio).
Here are some key points about metadata-based application roles:
• In general, roles do not protect data or metadata. Roles just control which features in a particular application are available to which users.
• An application feature that is under role-based management is called a capability.
Each role provides multiple capabilities. A user or group can be in multiple roles.
6
Chapter 1 • Security Overview
• Not all applications have roles. Not all application features are under role management. Each application that supports roles provides a fixed set of capabilities.
You cannot convert a feature that is not a capability into a capability.
T I P
If you add custom tasks (in SAS Enterprise Guide) or develop custom plugins (for SAS Management Console), you can register those features as capabilities. See
SAS Intelligence Platform: Desktop Application Administration
Guide
.
• Capabilities are additive. There are no negative capabilities (capabilities that limit what a user can do). It is not possible to deny a capability (capabilities are either granted or not granted).
• Capabilities can be categorized as follows: explicit capabilities can be incrementally added to or removed from any role (other than the unrestricted role, which always provides all explicit capabilities). Most roles have explicit capabilities.
implicit capabilities are permanently bound to a certain role. The metadata server's roles provide implicit capabilities. For example, the user administration role provides the capability to add users, but there is no explicit Create Users capability.
contributed capabilities are implicit or explicit capabilities that are assigned through role aggregation. If you designate one role as a contributing role for another role, all of the first role's capabilities become contributed capabilities for the second role.
• You cannot assign permissions to a role. You cannot assign capabilities to a group.
• A user cannot temporarily assume or relinquish a role. All of a user's roles are active at all times.
T I P
You can give an administrator two user definitions. This enables the administrator to function as a regular user some of the time.
• For details about a particular application's capabilities and roles, see the administrative documentation for that application.
For more information, see
.
Single Sign-On
The platform supports single sign-on (SSO) as follows:
• To bypass the initial logon prompt when launching a desktop application, use
Integrated Windows authentication.
• To bypass the initial logon prompt when launching a web application, configure web authentication.
• To avoid secondary logon prompts when accessing data servers and processing servers, use a combination SAS token authentication, Integrated Windows authentication, and credential management.
T I P
As a convenience for users who do not use SSO when they log on, you can
permit users to save their logon credentials on the client side. See “Client-Side
Storage of Passwords” on page 16
.
See Also
•
Chapter 10, “Authentication Model,” on page 113
•
Chapter 11, “Authentication Mechanisms,” on page 133
Single Sign-On
7
8
Chapter 1 • Security Overview
Chapter 2
User Administration
Availability of Application Features in a New Deployment . . . . . . . . . . . . . . . . . . . 16
Configuring Client Machines to Use Integrated Windows Authentication . . . . . .
9
10
Chapter 2 • User Administration
About User Administration
Introduction
In order to make access distinctions and track user activity, a security system must know who is making each request. In the platform, the primary user administration task is to store each user's external account ID in the SAS metadata. SAS uses its copy of these
IDs to establish a unique SAS identity for each connecting user. All of a user's metadatalayer memberships, permissions, and capabilities are ultimately tied to the user's SAS identity.
Note: It is not necessary to store passwords in the SAS metadata for the purpose of
identifying a user. SAS identity is determined by examining stored user IDs, not by examining stored passwords.
Note: For some service identities and metadata administrators, you can use a SAS
internal account instead of a stored SAS copy of an external account ID.
Who Can Manage Users, Groups, and Roles?
In the initial configuration for a new deployment, the SAS Administrators group has the user administration role, so members of that group can perform almost all user management tasks. The following table outlines the distribution of user administration capabilities.
Table 2.1 User Administration Capabilities
Metadata Server Role
Unrestricted
User administration
None
Actions Supported
Perform all identity management tasks.
Add, modify, and delete most identities.
Update your own personal logins.
For restricted user administrators (users who have the user administration role but are not unrestricted), the following constraints apply:
• Restricted user administrators cannot update the unrestricted role.
• To update or delete an identity, restricted user administrators must have the
WriteMetadata permission for that identity. For example, you can prevent a restricted user administrator from updating UserA’s metadata definition by taking away his or her default grant of the WriteMetadata permission (on UserA’s Authorization tab, explicitly deny the WriteMetadata permission to the restricted user administrator).
• To change a role's capabilities, restricted user administrators must have the
WriteMetadata permission for the associated software component.
• To access user management features in SAS Management Console, restricted user administrators must have the User Manager capability.
Metadata Users
11
Note: You can delegate administration of an existing identity to someone who is not a
user administrator. In the target identity's metadata definition, explicitly grant the
WriteMetadata permission to the delegated administrator.
Where is User Administration Performed?
Metadata-layer user administration is performed as follows:
• To manage identity information interactively, use SAS Management Console. See
SAS Management Console: Guide to Users and Permissions
.
• To import identity information in bulk from an external user store (such as Active
Directory) to the SAS metadata, write SAS code. See
.
• To copy identity metadata from one SAS repository to another, use the metadata promotion tools. See “Promotion Tools Overview” in SAS Intelligence Platform:
System Administration Guide
.
• To audit changes to metadata identity definitions, use the
Audit.Meta.Security.GrpAdm and Audit.Meta.Security.UserAdm log categories. See
See Also
•
“How SAS Identity Is Determined” on page 114
•
“PUBLIC Access and Anonymous Access” on page 130
Metadata Users
In general, each SAS user has identity information in two distinct realms:
1. In an authentication provider, the user has an account that can access the metadata server.
2. In the SAS metadata, the user has a definition that includes a copy of the account ID with which the user accesses the metadata server.
Coordination between these two realms establishes a unique SAS identity for each user.
Each SAS identity is based on a match between the following two values:
• the account ID with which the user authenticates
• the account ID that is listed in the user's metadata definition
In the following figure, account refers to a user account in an authentication provider, and definition refers to a metadata object that represents the user. Bill cannot log on,
Susan has only the generic PUBLIC identity, and Tara has an individual SAS identity.
12
Chapter 2 • User Administration
Figure 2.1 Examples: User Accounts and User Definitions
User account?
User account?
User definition?
User account?
User definition?
Bill
No SAS identity
Susan
PUBLIC
Generic SAS identity
Tara
Tara Individual SAS identity
ETL
SASUSERS
PUBLIC
Direct membership
Implicit memberships
Here are some tips for working with user definitions:
• If the metadata server runs on Windows and uses SAS authentication, the SAS copy of each user's Windows user ID must be stored in a fully qualified format (for example, WindowsDomain\user-ID, MachineName\user-ID, or user-
[email protected]).
• If you find that a user has only the PUBLIC identity even though the user has a user definition, the user's stored account ID might be missing, not accurately entered, or not in the correct format. Passwords and authentication domain assignments are never the cause of this problem. The match is based only on the account ID.
• Regular users (non-administrators) can maintain their own logins, but cannot make other changes to their definitions.
• Permission settings on a user definition do not determine what that user can do.
Those settings can affect the ability of other identities to update or delete the user definition itself. Special rules automatically protect user, group, and role definitions.
The following table introduces some of the predefined users:
Table 2.2 Selected Predefined Users
User
SAS Administrator
Description
A predefined super user that has unrestricted access in the metadata layer.
A predefined first user that can be useful for demonstrations.
SAS Demo User
SAS Trusted User
A service identity that can act on behalf of other users.
SAS Anonymous Web User
A service identity that can provide anonymous access for a few web components.
*
*
See Also
•
“About User Administration” on page 10
•
“Create Metadata User Definitions” on page 33
Metadata Groups
13
Metadata Groups
Groups are primarily used in access controls, because it is more efficient to assign permissions to groups than to individual users. You can also use a group to populate a role or to make a shared credential available to multiple users. The following figure illustrates how the users in the previous topic might participate in a group structure:
Figure 2.2 Example: Users in a Group Structure
Bill
PUBLIC
Susan
ETL
SASUSERS
SAS
Administrators
Tara
Report
Users
Here are some tips for working with group definitions:
• You can create a nested group structure by making one group a member of another group.
• Most groups do not have logins (stored credentials). A group login makes a shared external account available to all members of the group. Such outbound logins typically provide access to a third-party database server and should include both a user ID and a password (as well as an authentication domain).
• Permission settings on a group definition do not determine what that group can do.
Those settings can affect the ability of other identities to update or delete the group definition itself. Special rules automatically protect user, group, and role definitions.
The following table introduces some of the predefined groups.
Table 2.3 Selected Predefined Groups
Group
PUBLIC
SASUSERS
Description
Automatically includes everyone who can access the metadata server, either directly or through a trust relationship. A user who does not have an individual identity has only the PUBLIC group identity.
Automatically includes those members of the PUBLIC group who have an individual identity. All members of the SASUSERS group are also members of the PUBLIC group.
14
Chapter 2 • User Administration
Group
SAS Administrators
SAS General Servers
SAS System Services
Description
A standard group for metadata administrators. In a standard configuration, membership in this group provides broad access and most administrative capabilities, but does not provide unrestricted status.
A service group that enables its member (the SAS Trusted User) to see server launch credentials.
*
A service group that enables its member (the SAS Trusted User) to see servers, cubes, and other objects.
*
Server launch credentials (for example, the SAS Spawned Servers account, sassrv) are stored in logins on the Accounts tab of the SAS
General Servers group. This facilitates launching of stored process servers and pooled workspace servers.
See Also
•
“Identity Hierarchy” on page 17
•
“Differences between Roles and Groups” on page 15
Metadata Roles
Role Definitions
Roles control the availability of application features such as certain menu items, plugins, and buttons. In the initial configuration, registered users have almost all nonadministrative capabilities. In many cases, this is appropriate and sufficient. However, you can choose to alter the initial configuration by using either or both of the following techniques:
• To increase or reduce the availability of a role, change the assignment of members to the role.
• To redistribute capabilities, change the assignment of capabilities to roles, or create additional roles.
Here are some tips for working with roles:
• Before you make changes, make sure you have a complete and current backup.
• Capability assignments can be redundant across roles. To prevent someone from having a capability, you must make sure they are not in any role that provides that capability.
• Never change the name of a predefined role.
• There are no negative capabilities (capabilities that limit what someone can do). You cannot deny a capability to anyone.
• Some roles have implicit capabilities (capabilities that are bound to the role and are not displayed in the user interface).
• Contributing role relationships are monolithic. You cannot deselect a contributed capability. These relationships are also dynamic. A change to the capabilities of one role affects any roles to which the first role contributes its capabilities.
Metadata Roles
15
• Permission settings on a role definition do not determine what that role can do.
Those settings can affect the ability of other identities to update or delete the role definition itself. Special rules automatically protect user, group, and role definitions.
Predefined Roles
The predefined roles and capabilities are described in the administrative documentation for the associated application. The following table helps you locate documentation for predefined roles.
Table 2.4 Documentation of Predefined Roles
Application
Metadata Server
Desktop Applications
Web Applications
Visual Analytics: Basic
Visual Analytics: Report Viewing
Visual Analytics: Analysis
Visual Analytics: Data Building
Visual Analytics: Administration
Documentation of Predefined Roles
SAS Intelligence Platform: System
Administration Guide
SAS Intelligence Platform: Desktop
Application Administration Guide
SAS Intelligence Platform: Web Application
Administration Guide
SAS Visual Analytics: Administration Guide
SAS Visual Analytics: Administration Guide
SAS Visual Analytics: Administration Guide
SAS Visual Analytics: Administration Guide
SAS Visual Analytics: Administration Guide
Differences between Roles and Groups
Roles and groups serve distinct purposes. You cannot assign permissions to a role or capabilities to a group. Here are some additional distinctions:
• The identity hierarchy is relevant for groups, but not for roles. If you are a member of a role, you have all of that role's capabilities, regardless of whether you are a direct member of that role and what your other memberships are.
• You can deny a permission to a group, but you cannot deny a capability to a role.
Each role either provides or does not provide each capability. No role takes capabilities away from its members.
• A group's permissions are not displayed as part of a group definition, but a role's capabilities are displayed as part of a role definition.
• A group can be a member of another group, but a role cannot be a member of another role. Instead, one role can contribute its capabilities to another role.
16
Chapter 2 • User Administration
Availability of Application Features in a New Deployment
In general, the initial configuration provides appropriate access to application features.
Most nonadministrative capabilities are available to either PUBLIC (everyone who can access the metadata server) or SASUSERS (those members of PUBLIC who have a well-formed user definition). To ensure appropriate availability of features for your applications, see the administrative documentation for each application.
Password Management
Password Policies
Each authentication provider sets password policies for accounts in that provider. For example, the password expiration policy for a host account is determined by that host.
For the SAS internal authentication provider, you can set server-level policies (in the metadata server’s omaconfig.xml file) and per-account policies (in a user’s metadata definition). See
“How to Change Internal Account Policies” on page 177
.
Client-Side Storage of Passwords
In the initial configuration, users can choose to store their credentials in their client-side connection profiles. This prepopulates the logon dialog box in desktop applications.
For most desktop applications, the SASSEC_LOCAL_PW_SAVE= option controls the availability of a check box that enables users to choose whether to store credentials locally. To prevent users from creating a local copy of their credentials, set
SASSEC_LOCAL_PW_SAVE="N" (or ="0" or ="F") in the metadata server's omaconfig.xml file and restart the server.
Note: A change to the SASSEC_LOCAL_PW_SAVE= setting takes effect after the
metadata server is restarted. Each client uses the previous setting for its first connection, discovers the revised metadata server setting, and conforms to that revised setting for subsequent connections. If you change the setting to disallow saved credentials, and credentials are already present in a user's connection profile, those credentials must be manually removed.
Note: For a few solutions rich clients (for example, SAS Model Manager, SAS
Enterprise Miner, and SAS Forecast Studio), the ability to store credentials in clientside connection profiles is instead controlled by the
Policy.AllowClientPasswordStorage property. This property is available on the Plug-
ins tab in SAS Management Console (under Application Management
ð
Configuration Manager
ð
SAS Application Infrastructure
ð
Settings
ð
Policies) as the property Allow client password storage.
External Login Passwords
In most cases, the SAS copy of an external account includes only a user ID. For these cases, no password updates in metadata are necessary.
For any external passwords that are stored in the SAS metadata, updates are driven by changes that first occur in the external authentication provider. For example, if a copy of
Identity Hierarchy
17
the password for an Oracle account or a host account is stored in a group login, you must maintain that copy so that it always matches the actual password. Any change to the actual password (in Oracle) must be followed by a corresponding update to the SAS copy of the password (in the group login in the SAS metadata).
You can update stored passwords in SAS Management Console. If you own logins that include passwords, you can also update those passwords in SAS Personal Login
Manager. For example, to update the SAS copy of an external password in SAS
Management Console, navigate to the owning user or group definition, select the
Accounts tab, select a login, and click Edit.
Internal Account Passwords
Every SAS internal account has a password. By initial policy, these passwords do not expire.
To update a SAS internal password in SAS Management Console, navigate to the owning user definition, select the Accounts tab, and click Update (at the bottom of the tab). If you have your own SAS internal account, you can also update your internal password in SAS Personal Login Manager.
T I P
If repeated attempts to log on with an internal account fail, that account might be locked. See
“Unlock an Internal Account” on page 37 .
Managed Passwords
Passwords for a few service accounts require special coordination because these passwords are included in configuration files. To update these passwords, use the SAS
Deployment Manager.
See Also
•
“SAS Internal Authentication” on page 143
•
“Update a Managed Password” on page 35
Identity Hierarchy
The identity hierarchy can affect authorization decisions and login priority (in credential retrieval from the SAS metadata). The identity hierarchy is not relevant for roles.
The identity hierarchy establishes the following precedence ranking:
1. the user's individual identity, based on the user's authenticated ID.
2. a group that has the user as a direct member. This is a first-level group membership for the user.
3. a group that has another user group as a direct member. For example, assume that the user belongs to a group named ETL_Advanced, and that group is a member of another group called ETL_Basic. In that case, the ETL_Basic group is a second-level group for that user. If you have additional levels of nesting, each successive level has less precedence.
18
Chapter 2 • User Administration
4. the SASUSERS implicit group, which includes everyone who has an individual identity.
5. the PUBLIC implicit group, which includes everyone who can access the metadata server (regardless of whether they have an individual identity or not).
The following table provides examples of the hierarchy:
Table 2.5 Examples of Identity Hierarchies
Scenario
User has no individual identity.
User has an identity and no explicit group memberships.
User is a direct member of GroupA and GroupB.
GroupA is a member of the Report Users group.
User's Identity Hierarchy
Primary identity: PUBLIC
Primary identity: self
First-level memberships: SASUSERS
Second-level memberships: PUBLIC
Primary identity: self
First-level memberships: GroupA, GroupB
Second-level memberships: Report Users
Third-level memberships: SASUSERS
Fourth-level memberships: PUBLIC
T I P
To avoid introducing unnecessary complexity, do not make PUBLIC or
SASUSERS a member of another group. For example, if you make PUBLIC a member of GroupA, then a user who is an indirect member of GroupA (through his automatic membership in PUBLIC) has GroupA as his lowest precedence membership. This contradicts the usual expectation that every user's lowest precedence membership is PUBLIC. It is not a problem for PUBLIC or SASUSERS to be a member of a role.
Note: Avoid creating circular or reciprocal group memberships (for example, GroupA is
in GroupB, GroupB is in GroupC, and GroupC is in GroupA). Such relationships introduce unnecessary complexity and can negatively impact performance. When you define nested group memberships (interactively or through importing identity information), the best practice is to use as simple and non-redundant a structure as possible.
See Also
•
“Two Relationship Networks” on page 44
•
“Authorization Decisions” on page 60
External Identities
An external identity is a synchronization key that facilitates coordination between identity entries in the metadata and identity entries in your authentication provider. If you use batch processes to coordinate metadata identity information with your authentication provider, external identities are set up and used as follows:
Windows Privileges
19
1. In your authentication provider, you select a field to use for the mapping. This should be a field that contains a unique and unchanging value for each user, group, and role that you want to manage with batch processes. Typically, this is an identifier such as employee number.
2. When you perform an initial import from your authentication provider into the metadata, the keyid values in the canonical tables become external identity values in the metadata. Each imported identity has at least one external identity value.
3. During the synchronization process, external identity values that are extracted from the metadata are used as the keyid in the target tables. Because these values also exist in the extraction from your authentication provider, external identity values can be used to match corresponding entries in the two sets of tables.
Note: If you need to incorporate manually created identities into a batch synchronization
process, manage each identity’s external identity value from the General properties of its metadata definition.
See Also
Appendix 2, “User Import Macros,” on page 229
Windows Privileges
Access This Computer from the Network
Table 2.6 Access This Computer from the Network
Description
To Whom
How
This privilege is required in order to connect to SAS servers.
Give this privilege to all users who access SAS servers on Windows.
Typically, this right is already granted to the Windows group Everyone. To confirm, check the
Windows local policy settings.
Log on as a Batch Job
Table 2.7 Log on as a Batch Job
Description
To Whom
This privilege is required in order to run a stored process server or any type of workspace server.
On the Windows computer that hosts the SAS object spawner, give this privilege to the accounts under which workspace servers and stored process servers run:
• any service account under which one of these servers run
• all puddle logins for any client-side pooled workspace servers
• any user accounts under which a standard workspace server runs (users who authenticate by
Integrated Windows authentication or SAS token authentication do not need this privilege)
20
Chapter 2 • User Administration
How Modify the local security policy. For example, on Windows XP, this right is managed from the
Windows control panel under Administrative Tools ð Local Security Policy ð User Rights
Assignment ð Log on as a batch job. If you have an operating system group (such as SAS Server
Users) that has this right, you just add users and service account identities to that group.
Trusted for Delegation
Table 2.8 Trusted for Delegation
Description
To Whom
How
This privilege enables a process to allow each user's credentials to be sent to further machines for authentication (for example, to access a UNC path). The privilege is needed if the workspace server is accessed through Integrated Windows authentication and provides access to network resources.
Note: With Integrated Windows authentication, the workspace server does not receive the requesting user's credentials, so the workspace server cannot provide credentials for downstream servers. Instead, the spawner account must be trusted to delegate each requesting user's identity as necessary.
If the workspace server runs on Windows, give this privilege to the account under which the object spawner runs. By default, the spawner runs as a service under the local system account, so the computer account for spawner's host needs the privilege.
If the workspace server runs on UNIX, give this privilege to the service principal account that is referenced in the relevant keytab (the keytab is based on service principal names that correspond to a particular service principal account). For more information, see the chapter "Configuring Integrated
Windows Authentication" in Configuration Guide for SAS Foundation for UNIX Environments at http://support.sas.com/documentation/installcenter ).
As a Windows domain administrator, under Start ð Control Panel ð Administrative Tools ð
Active Directory Users and Computers, access the properties dialog box for the relevant account and grant the privilege.
If your spawner runs on Windows under the local system account, select the spawner host machine in
Active Directory under Computers. On the Delegation tab (or the General tab), select the Trust this
computer for delegation check box.
If your spawner runs on Windows under a domain account, select that account in Active Directory under Users. On the Delegation tab (or the Accounts tab), select the Account is trusted for
delegation check box.
If your spawner runs on UNIX, select the appropriate service principal account in Active Directory under Users. On the Delegation tab (or the Accounts tab), select the Account is trusted for
delegation check box.
Note: In most cases, an object spawner on Windows runs as a service under the local
system account. If the spawner instead runs under some other account, that account must be a Windows administrator on the spawner's host and have the Windows user rights Adjust memory quotas for a process and Replace a process level token.
These user rights assignments are part of the local security policy for the Windows computer that hosts the spawner.
See Also
•
“Host Authentication” on page 137
•
“Integrated Windows Authentication” on page 139
Windows User ID Formats
21
Windows User ID Formats
In most cases, users can launch SAS applications using the same ID and password as they use in the rest of your computing environment. However, when you create a SAS copy of a Windows user ID, you must store the user ID in the appropriate format. In most cases, you must use a particular qualified format (for example, WindowsDomain
\user, MachineName\user, or [email protected]). With certain authentication
configurations, a different format is required. Failure to appropriately qualify a stored user ID causes the user to have only the PUBLIC identity.
If your site accepts Windows IDs in disparate formats, you must coordinate the format of the copies with the format in which users submit their IDs. This table describes the common forms for an Active Directory user ID:
Table 2.9 Overview: Forms of an Active Directory User ID
Form
Short
UPN
Basic Syntax
user user@UPNsuffix
Down-level
down-level-domain-name\user
Kerberos
user@REALM
Examples joe [email protected]
or
orionsports\joe
or
sales\joe
or
mymachine\joe [email protected]
Note: User Principal Name (UPN) is an Active Directory concept. Down-level domain
is a Windows NT concept. The realm in a Kerberos principal is usually a Windows domain. A Kerberos principal can include an instance (in the format user/
instance@REALM). Additional site-specific variations might occur. The realm
qualifier in a Kerberos principal is always in uppercase.
In the platform, follow these guidelines for Windows user IDs:
• If users log on interactively, they can usually use the short form. Here are some exceptions:
• The user has multiple accounts with the same user ID in different down-level domains (for example,
machine\joe
,
domain1\joe
, and
domain2\joe
).
• The site has configured direct use of LDAP and has not specified
-primpd
(the
PRIMARYPROVIDERDOMAIN system option).
• If users log on interactively, they can also use one other site-supported form (either the UPN form or the down-level form). Use one of these approaches:
• In the metadata, store each user ID in UPN form. Tell users not to use the downlevel form when they log on.
• In the metadata, store each user ID in down-level form. Tell users to not use the
UPN form when they log on.
• If users log on to SAS desktop applications through Integrated Windows authentication, their user IDs should usually be stored in down-level form. In
22
Chapter 2 • User Administration
general, that is the form in which SAS obtains user IDs after Kerberos authentication occurs.
Note: If you prefer to store user IDs in the native Kerberos form, add the setting
SASUSEKERBNAME true
as a Windows system environment variable on the server host. For example, on the Windows desktop, right-click My Computer, select Properties, select the Advanced tab, click the Environment Variables button, add the setting under System variables, and reboot the machine. This setting affects only connections that use Integrated Windows authentication. If you use this setting, you might want to make sure that the Integrated Windows authentication process always chooses the Kerberos protocol.
• If users log on to SAS web applications through Integrated Windows authentication
(which occurs only if you configure web authentication and have set up Integrated
Windows authentication with your web provider), the form of the returned user ID might differ. See the documentation for your web application server.
Note: In the status bar of some applications, a currently connected Windows user ID is
always displayed in the format user@VALUE, regardless of how the user logged on or how the user's ID is stored in the metadata. For example, if you log on as Joe and your stored user ID is WIN\joe, the status bar displays your authenticated ID as
joe@WIN.
See Also
•
“Direct LDAP Authentication” on page 136
•
“How SAS Identity Is Determined” on page 114
•
“Integrated Windows Authentication” on page 139
•
“Authentication to the Metadata Server” on page 114
Uniqueness Requirements
The metadata server enforces the following identity-related constraints:
• You cannot create a user definition that has the same name as an existing user definition. The display names do not have to be unique.
T I P
We recommend that you avoid using spaces or special characters in the name of a user, group, or role. Not all components support spaces and special characters in identity names.
• You cannot change the name of an existing user, group, or role in SAS Management
Console. You can change the display name.
• You cannot create a group or role definition that has the same name as an existing group or role definition. The display names do not have to be unique.
• You cannot assign the same external account ID to two different identities. All of the logins that include a particular ID must be owned by the same identity. This requirement enables the metadata server to resolve each ID to a single identity.
However, a login that is associated with an outbound authentication domain is exempt from this requirement. For more information, see
Authentication Domains” on page 124 .
See Also
Configuring Client Machines to Use Integrated Windows Authentication
23
• This requirement is case-insensitive. For example, you cannot assign a login with an ID of smith to one user and a login with an ID of SMITH to another user.
• This requirement applies to the fully qualified form of the ID. For example, you can assign a login with an ID of winDEV\brown to one user and a login with an
ID of winPROD\brown to another user. In this example, winDEV and winPROD are Windows domain names, which are incorporated into the fully qualified form of an external account ID.
• This requirement cannot be mitigated by associating the logins with different
SAS authentication domains. For example, assume that one user has a login with an ID of smith that is associated with a SAS authentication domain named
DefaultAuth. In that case, you cannot give any other user a login with the ID
smith, even if you plan to assign the login to a different SAS authentication
domain.
T I P
To enable multiple users to share an account, store the credentials for that account in a login as part of a group definition. Then add the users who share the account as members of that group definition.
• If you give a user two logins that contain the same ID, the logins must be associated with different authentication domains. Within an authentication domain, each ID must be unique. For example, if you give the person Tara O'Toole two logins that both have an ID of tara, then you cannot associate both of those logins with the
OraAuth authentication domain.
Note: Like the previous requirement, this requirement is case-insensitive.
Note: You cannot circumvent this constraint by supplying the same user ID in
different formats. For example, if you give a user a login that contains the user
ID WIN\user and another login that contains the user ID user@WIN, those logins must be in different authentication domains.
• A user can have multiple locations, email addresses, and telephone numbers.
However, each user can have only one item of a given type. For example, a user can have one home email address and one work email address, but not two work email addresses.
•
“Create Metadata User Definitions” on page 33
•
“How SAS Identity Is Determined” on page 114
Configuring Client Machines to Use Integrated
Windows Authentication
About Configuring Client Machines to Use IWA
This information applies to client machines that are running Java rich clients, including those that might run clients launched through Java Web Start. In addition to enabling the
Use Integrated Windows authentication (single sign-on) option during installation, there are manual steps that must be completed from each client machine. The following sections provide information about the manual steps that must be completed.
24
Chapter 2 • User Administration
Updating Registry Settings
By default, Windows does not allow Java to access the session key that is associated with the ticket-granting ticket (TGT) in the cache. This might cause the Java client to prompt the user for their password, or it might refuse to authenticate with the server. To enable Java access to the session key, complete one of the following workarounds:
• Update the Windows registry. Microsoft recommends performing a system backup before editing the registry. Access the HKEY_LOCAL_MACHINE\System
\CurrentControlSet\Control\Lsa\Kerberos\Parameters registry subkey. Add the
DWORD value with a name of allowtgtsessionkey and value of 1. For more information, see http://support.microsoft.com/kb/308339 .
• Run the
kinit -f
command to get a new TGT and session key from the Key
Distribution Center and cache it on the filesystem.
Note: The TGT has a limited lifespan. If you chose this workaround, you have to
periodically repeat this process.
Note: Use the kinit command that is provided in the
SASHOME
\SASPrivateJavaRuntimeEnvironment\9.4\jre\bin
directory.
Configuring the Java Cryptography Extension
Determining Whether a Machine Is Using AES-256
IWA requires that the Java environment be updated with the Java Cryptography
Extension (JCE) Unlimited Strength Jurisdiction Policy files if the machine is using
AES-256 encrypted tickets or session keys. By default, Java is not shipped with support for AES-256 enabled. Kerberos uses AES-256 in the aes256-cts-hmac-sha1-96 encryption type that is associated with the keytab.
Determine whether a machine is using AES-256 encrypted TGT by running the
klist
command. The command displays currently cached Kerberos tickets. Here is an example of the output that you might see:
C:\Users\user_name>klist
Current LogonId is LUID
Cached Tickets: (number)
#0> Client: user_name @ domain
Server: krbtgt/domain @ domain
KerbTicket Encryption Type: AES-256-CTS-HMAC-SHA1-96
Ticket Flags 0x60a00000 -> forwardable forwarded renewable pre_authent
Start Time: 5/12/2014 11:09:03 (local)
End Time: 5/12/2014 19:50:18 (local)
Renew Time: 5/19/2014 9:50:18 (local)
Session Key Type: AES-256-CTS-HMAC-SHA1-96
#1> Client: user_name @ domain
Server: krbtgt/domain @ domain
KerbTicket Encryption Type: AES-256-CTS-HMAC-SHA1-96
Ticket Flags 0x40e00000 -> forwardable renewable initial pre_authent
See Also
Configuring Client Machines to Use Integrated Windows Authentication
25
Start Time: 5/12/2014 9:50:18 (local)
End Time: 5/12/2014 19:50:18 (local)
Renew Time: 5/19/2014 9:50:18 (local)
Session Key Type: AES-256-CTS-HMAC-SHA1-96
Note: Use the klist command that is provided in the
SASHOME
\SASPrivateJavaRuntimeEnvironment\9.4\jre\bin
directory.
Enabling AES-256
To enable AES-256, you must download the Unlimited Strength Java Cryptography
Extension (JCE) policy files and install them into the directory where the JRE is configured, follow these steps:
1. Download the JCE Unlimited Strength Jurisdiction Policy files from Oracle. Be sure to download the version that corresponds to the version of Java that is installed. It is available from http://www.oracle.com/technetwork/java/javase/downloads/ index.html
.
2. Extract the archive. In the
jce
directory, extract all three files (local_policy.jar,
README.txt, and US_export_policy.jar) to
JAVA_HOME\jre\lib\security
.
Note: For a SAS installation, the default JRE is located in
SASHome
\SASPrivateJavaRuntimeEnvironment\9.4\jre
. For client machines using Java Web Start, the JRE location depends on the Java Web Start configuration. Verify the JRE settings in the Java Control Path by navigating to
Control Panel
ð
Java.
•
“Integrated Windows Authentication” on page 139
•
“How to Configure Integrated Windows Authentication” on page 165
26
Chapter 2 • User Administration
27
Chapter 3
Access Management
WriteMetadata and WriteMemberMetadata . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
About Access Management
Introduction
The platform provides a metadata-based authorization layer that supplements protections from the host environment and other systems. You can use the metadata authorization layer to manage access to almost any metadata object (for example, reports, data definitions, information maps, jobs, stored processes, and server definitions). For information specific to SAS Visual Analytics, see SAS Visual Analytics: Administration
Guide.
Across authorization layers, protections are cumulative. In order to perform a task, a user must have sufficient access in all applicable layers.
In the metadata layer, the following permissions are always enforced:
• the ReadMetadata permission (RM), which controls the ability to see an object
• the WriteMetadata permission (WM), which controls the ability to update or delete an object
Other permissions are specialized and affect only certain types of objects.
CAUTION:
28
Chapter 3 • Access Management
In the metadata authorization layer, not all permissions are enforced for all items.
It is essential to understand which actions are controlled by each permission.
See “Use and Enforcement of Each Permission” on page 67 .
CAUTION:
Some clients enable power users to create and run SAS programs that access data directly, bypassing metadata-layer controls.
It is essential to manage
Who Can Set Permissions?
Table 3.1 Requirements for Setting Permissions
Task
Set permissions on an item
Change the permission pattern on an ACT
Designate a different repository ACT
Requirements
WriteMetadata for the item
WriteMetadata for the ACT
WriteMetadata for the ACT
Note: In SAS Management Console, you cannot see the Authorization Manager or
any Authorization tabs unless you have the Authorization Manager capability.
Where is Access Management Performed?
Metadata-layer access management is performed as follows:
• To manage access interactively, use SAS Management Console (a desktop application). See
SAS Management Console: Guide to Users and Permissions
.
• To manage metadata access, use SAS Environment Manager (a web-based solution).
See SAS Environment Manager: User’s Guide.
• To programmatically define or query authorization settings, use either batch tools
(see
Chapter 9, “Batch Tools for Metadata Authorization,” on page 103 ) or DATA
step functions (see
SAS Language Interfaces to Metadata
).
• To audit changes to access controls, use the Audit.Meta.Security.AccCtrlAdm log category. See
.
• To extract authorization information from the SAS metadata into SAS data sets, use the %MDSECDS macro. See
“Overview of the Security Report Macros” on page
• To copy metadata access controls from one SAS repository to another, use the metadata promotion tools. See “Promotion Tools Overview” in SAS Intelligence
Platform: System Administration Guide
.
Basics of Metadata Authorization
29
Basics of Metadata Authorization
Where Permissions Are Set
Effective access to an object is displayed in the authorization properties section of the object's metadata definition. You can set permissions on individual content objects (such as reports) and on containers (such as folders). For simplicity, try to set permissions on containers, instead of on individual content objects, whenever possible.
T I P
The metadata authorization model is object-centric, not identity-centric. To examine a user's permissions, do not begin by finding their user definition. Instead, begin by navigating to the object that you want to examine.
Who Permissions Are Assigned To
You can assign permissions to individual users and to groups. For simplicity, try to avoid assigning permissions to individual users.
The main authorization list includes only those users and groups that participate in access controls that could affect the object. The list usually includes at least the following groups:
PUBLIC automatically includes everyone who can access the metadata server.
SASUSERS automatically includes everyone who is registered in the metadata (SASUSERS is a subset of PUBLIC).
SAS Administrators should include only metadata administrators that require broad access.
SAS System Services should include only the service identity (or, identities) that require broad access.
Anyone who is not listed in the basic authorization properties display has the access of their closest listed group, as determined by group memberships and identity precedence.
Here are some examples:
• The closest (and only) listed group for an unregistered user is PUBLIC.
• The closest listed group for a registered user is often SASUSERS.
• The closest listed group for an administrator is usually SAS Administrators.
How Permissions Are Set
Use a combination of the following techniques to manage access: direct controls
You can assign grants and denials directly on a target object. There are two types of direct controls: explicit controls are individual grants and denials.
30
Chapter 3 • Access Management
access control template (ACT) controls are grants and denials within a predefined pattern.
T I P
For simplicity, try to avoid setting unnecessary (redundant) direct controls.
object inheritance
You can assign grants and denials on a parent object, and rely on inherited settings to protect each child object. For example, a report inherits effective permissions from its parent folder. In the metadata layer, access control inheritance happens automatically and cannot be turned off. All objects have inherited settings. Most objects have inherited settings only.
The following table summarizes the three types of access controls:
Table 3.2 Direct Access Controls and Indirect Settings
Term
Direct control: explicit
Direct control:
ACT
Indirect setting
Significance
An access control is explicit for a particular identity if the control is set directly on the target object and assigned directly to that identity. For example, if an individual access control on ReportA grants the
ReadMetadata permission to UserA, then we say that UserA has a direct control on ReportA (specifically, an explicit grant of the ReadMetadata permission). Explicit controls are sometimes referred to as ACEs (access control entries).
An access control is a direct ACT control for an identity if the control comes from an ACT that is directly applied to the target object, with a pattern that directly assigns a grant or denial to that identity. For example, if the DemoACT is directly applied to FolderA, and the DemoACT’s permission pattern explicitly grants the ReadMetadata permission to
GroupA, then we say that GroupA has a direct control on FolderA
(specifically, an ACT-based grant of the ReadMetadata permission).
An indirect setting comes from someone else (a group that has an explicit or ACT setting), from somewhere else (a parent object, the repository
ACT), or from a special status (such as unrestricted). For the
WriteMemberMetadata permission, indirect means that the setting mirrors the WriteMetadata setting.
Note: The authorization model includes two distinct relationship networks: object inheritance (for example, a folder and its contents) and identity precedence (for example, a group and its members). A setting that is derived through either of these hierarchies is called an indirect setting.
How Permissions Differ from Capabilities
Having a certain capability is not an alternative to meeting permission requirements.
Permission requirements and capability requirements are cumulative. For example, even if you are in the Enterprise Guide: OLAP role, you cannot access data in a cube for which you do not have the Read and ReadMetadata permissions.
Use roles and permissions in conjunction with one another. You can use permissions to constrain the scope of a role. For example, to allow someone to create reports but add them to only one folder, give the person the Web Report Studio: Report Creation role, but grant them the WriteMemberMetadata permission on only that one folder.
See Also
Review: Key Points about Authorization
31
•
“Identity Hierarchy” on page 17
•
“Object Inheritance” on page 46
•
“Authorization Decisions” on page 60
•
“Use and Enforcement of Each Permission” on page 67
WriteMetadata and WriteMemberMetadata
You can use the WriteMetadata (WM) and WriteMemberMetadata (WMM) permissions to enable someone to interact with a folder's contents, but not update or delete the folder itself. On the folder, grant WMM and deny WM.
The following list explains the difference between these two permissions.
WriteMetadata (WM) provides general control for additions, deletions, and updates. For example, to edit a report, you need WM for the report. To delete a report, you need WM for the report
(and WMM for the report's parent folder). For containers other than folders (such as repositories, libraries, and schemas), WM also affects adding and deleting child objects. For example, to add an object anywhere in a repository, you need WM at the repository level. For folders, adding and deleting child objects is controlled by
WMM, not WM.
WriteMemberMetadata (WMM) provides specialized control for adding and removing objects in a folder. You need
WMM on a folder in order to add an object to the folder or delete an object from the folder. For example, to save a report to a folder, you need WMM for the folder. To remove a report from a folder, you need WMM for the folder (and WM for the report).
T I P
We recommend that anyone who has a grant of WM is not denied WMM.
T I P
If WMM is not directly set on a folder, the WMM setting mirrors the WM setting. WMM is never inherited from a parent object.
See Also
“WriteMetadata and WriteMemberMetadata” in SAS Management Console: Guide to
Users and Permissions
Review: Key Points about Authorization
In order to effectively use the metadata authorization layer, you must understand the following points:
• Metadata layer permissions supplement protections from the host environment and other systems. Across authorization layers, protections are cumulative. In order to perform a task, a user must have sufficient access in all applicable layers.
32
Chapter 3 • Access Management
• The displayed effective permissions are a calculation of the net impact of all applicable permission settings in the metadata layer. However, the display does not reflect access in other layers such as the operating system.
• Setting permissions is an object-centric activity. To define permissions for someone, do not begin by finding that person's user definition. Instead, begin by navigating to an object that you want to protect or make available.
• Explicit and ACT controls on an object (such as a report) always have priority over controls on the object's parent (such as a folder). For example, if a report has an explicit denial for PUBLIC and the report's folder has an explicit grant for you, the result is a denial for you. The direct denial for PUBLIC blocks access to the report
(for all restricted users), unless you offset the direct denial by adding direct grants on the report.
• In a user definition, the displayed permission settings have no effect on what the user can see or do. Those controls determine who can update or delete the user definition itself.
• Only the ReadMetadata and WriteMetadata permissions are universally relevant and enforced.
• The authorization layer that SAS provides supports very specific settings ( for example, on ReportA, grant the ReadMetadata permission to UserA). However, for simplicity, it is preferable to set permissions at a less granular level whenever possible. The recommended approach is to create a folder structure and set permissions so that each folder offers appropriate group-level access to its contents
(for example, on FolderA, grant the ReadMetadata permission to GroupA).
• After you add a broad denial, review the impact that this has on everyone else. For example, if the only setting on an object is an explicit PUBLIC denial, that denial blocks access for everyone (other than unrestricted users). To offset the denial, add one or more selective explicit (or ACT) grants.
• Before you deny the ReadMetadata permission on a folder, consider the navigational consequences. Without ReadMetadata permission on a folder, you cannot navigate to objects beneath that folder. Users need a clear path of grants of ReadMetadata permission in order to navigate to the content that they use.
See Also
•
“Adjust Initial Access” on page 38
•
“Use and Enforcement of Each Permission” on page 67
33
Chapter 4
Selected Tasks
About Security Task Instructions
This chapter provides information about selected security tasks.
For specialized configurations, see
Chapter 14, “Encryption Tasks,” on page 197 and
Chapter 12, “Authentication Tasks,” on page 155
.
For other user management and access management tasks, see
SAS Management
Console: Guide to Users and Permissions
.
Create Metadata User Definitions
Add Regular Users
For accountability, we recommend that you create an individual SAS identity for each person who uses the SAS environment. This enables you to make access distinctions and audit individual actions in the metadata layer. This also provides a personal folder for each user. To create a SAS identity for someone, store a copy of the person’s user ID in the SAS metadata.
Each user should have at least the following attributes:
• a name that is unique among users within the metadata server.
34
Chapter 4 • Selected Tasks
• a login that includes the user's external account ID. This might be any type of account that is known to the metadata server's host (an LDAP, Active Directory, host, or other type of account).
Note: For a Windows account, specify the user ID in an appropriate format. See
“Windows User ID Formats” on page 21 .
Here are some details and tips:
• As an alternative to adding users interactively, you can batch import users from a provider such as LDAP into the SAS metadata.
• The metadata server maintains its own copy of each ID, but does not keep copies of passwords for identification purposes.
• Registered users automatically belong to PUBLIC (everyone who can access the metadata server) and SASUSERS (those members of PUBLIC who have a wellformed user definition).
• A user who does not have a well-formed definition has only the PUBLIC identity. In the standard configuration, a PUBLIC-only user cannot access any resources. Not all applications allow a PUBLIC-only user to log on.
• If the workspace server is on Windows, give anyone who accesses that server using credential-based host authentication the Windows privilege Log on as a batch job.
• Authorization settings within a user definition do not determine what that user can do. Those settings affect the ability of other users to update or delete that user definition.
Add Administrators
For accountability, we recommend that you establish individual metadata administrators rather than sharing the unrestricted SAS Administrator account.
Each administrator should have at least the following attributes:
• a name that is unique among users within the metadata server
• either a login (as explained in the preceding topic) or an internal account
• membership in the SAS Administrators group
Here are some details:
• We recommend that you use an administrative identity only to perform tasks that require a high level of privilege.
• If you log on to SAS Management Console using an internal account, you are prompted for host credentials for certain tasks (for example, validating a standard workspace server).
• If you log on with an internal account, you must include the @saspw suffix in the user ID that you submit (for example, sasadm@saspw).
Add Dual Users
To enable someone to alternately function as an administrator and as a nonadministrator, create two user definitions for that person as follows:
• One definition is based on an internal account and is a member of the SAS
Administrators group.
See Also
Update a Managed Password
35
• The other definition is based on an external account and is not a member of the SAS
Administrators group.
Here are some tips for working with a dual user:
• The only way to make someone a dual user is to give that person two user definitions, each based on a different account. You cannot create a dual user by adding a login to a definition that already has an internal account or by adding two logins to one definition.
• A dual user should use a dedicated client-side connection profile for their internal account. In that profile, the user should leave the Authentication Domain field blank. This optimizes credential reuse.
• A dual user should log on with their internal account when they need administrative privileges and with their external account the rest of the time.
•
•
Appendix 2, “User Import Macros,” on page 229
•
“Windows Privileges” on page 19
•
“Windows User ID Formats” on page 21
•
“Identify or Create User Accounts” on page 156
•
“Logins for Users Who Participate in Web Authentication” on page 159
Update a Managed Password
Passwords for a few service accounts require special coordination because these passwords are included in configuration files. To update these passwords, use the SAS
Deployment Manager. Here are some key points about using the SAS Deployment
Manager to update passwords:
• The SharedServices password is required to change the password for sasadm and
SharedServices.
• The utility updates both configuration files and metadata. You can update multiple passwords in a single pass.
• You must run the utility on each machine that hosts affected components. If you have servers on multiple machines, run the utility on each host, beginning with the metadata server machine.
• It might be necessary to update the same password on multiple hosts. For example, if you update the password for the SAS Trusted User on the metadata server's host, you must also do the same update on the middle-tier machine.
• Be sure to supply the same new password for an account on all machines on which you update that account.
• If you enter a plaintext password into the utility, the utility encodes that password using SAS proprietary encoding (SAS002).
• Passwords for any service accounts that you introduce in SAS Management Console are not managed by this tool. For example, if you designate a new login as the launch
36
Chapter 4 • Selected Tasks
credential for a server, that launch credential is not automatically added to the list of accounts that the SAS Deployment Manager can update.
Note: Server launch credentials are not added to any configuration file, so you can
update any such passwords from the owning identity's Accounts tab in SAS
Management Console.
• You can automate running the deployment manager when you need to perform the same configuration action on many machines in your deployment. The deployment manager uses the same record and playback mechanism as the SAS Deployment
Wizard to perform a non-interactive, silent configuration.
CAUTION:
If you choose to use the deployment manager's record and playback mechanism to update passwords, passwords are written to the response file.
For greater security, delete the response file (or remove the passwords from the response file) when you are finished. A response file is present only if you use the record and playback mechanism, instead of completing the task manually as documented in the preceding steps.
• Each run of this utility generates an UpdatePasswords.html file that documents the updates that the utility performed and provides instructions for any required postupdate activities. The file is located in your equivalent of
\SAS\Config
\Lev1\Documents\
.
To update a password with SAS Deployment Manager:
Note: The procedure to update the SAS Environment Manager identity password is
different from the process detailed here. For more information, see SAS Environment
Manager: User’s Guide.
1. (Optional) If you are updating the password for an internal account, review the server-level password policies for internal accounts. Also, check each internal account's properties to determine whether any more (or less) stringent requirements apply.
Note: In particular, make sure that the account is not subject to a forced password
change after the password is reset (either set the password to never expire or change the server-level policy for pre-expired passwords).
Note: By default policy, internal passwords must be at least six characters and do
not have to include mixed case or numbers. The five most recent passwords for an account cannot be reused for that account.
2. (Optional) If you want to use stronger encoding than SAS002 (SASProprietary), use the PWENCODE procedure. For example, to prepare an AES-encrypted version of a new password, use the following code: proc pwencode in='PWsassrv1' method=sas003; run;
The encoded password is written to your SAS log. When you use method=sas003, the first part of the password is {SAS003}.
3. Stop all SAS servers and services.
4. If you are updating the password for an external account (for example, sassrv), change that password in your external authentication provider (for example, in the host operating system).
5. Restart the metadata server, the SAS Web Infrastructure database (usually the SAS
Web Infrastructure Platform Data Server), and any SAS solution-specific data servers. Do not restart other servers or services.
Unlock an Internal Account
37
6. On the metadata server’s host, navigate to your equivalent of
SAS-
installation-directory/SASDeploymentManager/9.4/
and launch sasdm.exe (Windows), sasdm.sh (UNIX), or sasdm.rexx (z/OS).
Windows Specifics
On Windows, log on as a Windows administrator of the current machine.
Operating Environment Information
On UNIX and z/OS, log on with the user ID that was used to perform the SAS installation.
7. In the SAS Deployment Manager, select the Update Passwords task, select a configuration directory on the current machine, and log on as an unrestricted user
(for example, sasadm@saspw).
8. Perform the update. If you need detailed assistance with the user interface, see the
Help within the utility.
9. If you have servers on multiple machines, repeat steps 6–8 on each server host as applicable for the accounts that you are updating. Remember that you might have to update the same account on multiple hosts.
Note: Not all accounts are used on all hosts. If the accounts that you are updating are
not on a particular host, proceed to the next host.
10. Restart all servers and services, and complete any additional post-update tasks as specified in the generated UpdatePasswords.html file.
Note: Because of dependencies, it is important to start servers and services in a
particular order. For a complete discussion, see the chapter "Operating Your
Servers" in
SAS Intelligence Platform: System Administration Guide
.
Unlock an Internal Account
By initial policy, making three consecutive unsuccessful attempts to log on with a SAS internal account causes that account to be locked for one hour. This topic explains how to unlock the account immediately, so that you do not have to wait until the account lockout period has passed.
The preferred approach is to locate another user who has user administration capabilities.
That user can unlock the internal account by completing these steps:
1. Log on to SAS Management Console as someone who is unrestricted or who has user administration capabilities.
2. On the Plug-ins tab, select User Manager.
3. In the display pane, clear the Show Groups and Show Roles check boxes. Rightclick the user definition of the person whose SAS internal account is locked out and select Properties.
4. Select the Accounts tab. A message box asks whether you want to unlock the account. Click Yes.
5. In the user's Properties dialog box, click OK. The account is now unlocked.
Note: It is not necessary to reset the user's internal password as part of unlocking the
user's internal account.
38
Chapter 4 • Selected Tasks
If there is no other administrator available and you have the necessary host access, you can use this approach as a last resort:
1. Edit the metadata server's adminUsers.txt file to create a new unrestricted user. a. Navigate to your equivalent of
SAS/Lev1/SASMeta/MetadataServer/ adminUsers.txt
. Open the file with a text editor.
b. Add a user ID for an account that is known to the metadata server's host. Include a preceding asterisk (for example,
*WIN\winID
or
*unixID
).
c. Stop and restart the metadata server to make the change take effect.
CAUTION:
Stopping the metadata server stops other components.
2. Log on to SAS Management Console using the user ID that you added to the adminUsers.txt file. In the status bar at the bottom of the application window, notice that you are unrestricted. Unlock the locked account. See the preceding instruction list for more information.
3. To verify that the account is unlocked, log on to SAS Management Console using the account.
4. Open the adminUsers.txt file, remove the entry that you added, and stop and restart the metadata server.
Here are some additional tips:
• If you choose to change the password for the original SAS Administrator, you might need to update the deployment.
• You can customize the account lockout policy.
• We recommend that you establish individual metadata administrators rather than sharing the predefined SAS Administrator account.
See Also
•
“SAS Internal Authentication” on page 143
•
“How to Change Internal Account Policies” on page 177
Adjust Initial Access
The initial configuration in a new deployment provides sufficient access to data and resources, with the following exceptions:
• Only unrestricted users can access data through information maps, reports that are based on information maps, the metadata LIBNAME engine, or the OLAP server. In the initial configuration, the only grants of the Read permission are in each user's personal content area (My Folder).
• Only unrestricted users and members of the SAS Administrators group can register cubes.
To ensure appropriate access to resources and data:
1. Log on to SAS Management Console as an administrator (for example, sasadm@saspw).
Adjust Initial Access
39
2. (Optional) Verify that all registered users have at least the minimum required repository-level access. a. On the Plug-ins tab, under Authorization Manager, expand the Access Control
Templates node.
b. Right-click the repository ACT (Default ACT) and select Properties. c. On the Permission Pattern tab, select SASUSERS. Verify that the
ReadMetadata and WriteMetadata permissions are granted.
3. (Optional) Verify that all registered users have basic access to the folder tree. a. On the Folders tab, right-click the root folder ( SAS Folders) and select
Properties. b. On the folder's Authorization tab, select SASUSERS. Verify that the
ReadMetadata permission is granted.
4. Provide metadata layer access to data (this is a broad approach). a. On the Authorization tab for the root folder ( SAS Folders), select
SASUSERS.
Note: To access this tab, select the Folders tab, right-click the root folder, and
select Properties.
b. Grant the Read permission. This enables registered users to perform tasks such as querying cubes, accessing data through information maps, and viewing the contents of tables.
If you want to manage access to data more precisely, set grants of the Read permission on specific folders for specific users. Users need the Read permission as follows:
• Users need the Read permission on an information map in order to access data through that information map. For example, if Joe is denied Read permission on an information map, he cannot view reports that are based on that information map.
• Users always need the Read permission on OLAP data in order to access that data.
• Users sometimes need the Read permission on relational data in order to access that data. Read permission is required when data is accessed using the metadata
LIBNAME engine.
• Users need the Select permission on metadata-bound data in order to access that
data. See “Metadata-Bound Libraries” on page 94 .
5. If users who are not in the SAS Administrators group will register cubes, grant those users the WriteMetadata permission on the OLAP schema. a. On the Folders tab, expand the
Shared Data
folder and select the SASApp -
OLAP Schema folder.
b. In the right panel, right-click the schema and select Properties. c. On the Authorization tab, select or add an identity and grant WriteMetadata permission to that identity. For example, to allow all registered users to add cubes, assign the grant of WriteMetadata permission to SASUSERS.
40
Chapter 4 • Selected Tasks
6. Verify that physical-layer access is available. Here are the general requirements:
• A user who accesses SAS data sets from a standard workspace server needs host layer Read access to those files.
• A user who performs tasks that involve writing to a host directory needs host layer Write access to that directory.
• Server launch credentials need host access to any SAS data that the server retrieves. Initially, the SAS Spawned Servers account (sassrv) is the launch credential for the stored process server and the pooled workspace server.
7. In the initial configuration, the Server Manager capability is available to only the
SAS Administrators group. This prevents other users from accessing server definitions under that plug-in. For greater security, use permissions to protect server
definitions. See “Protect Server Definitions” on page 81 .
8. In a new deployment, access to most resources and data is undifferentiated. All registered nonadministrators have identical metadata-layer access to content, data, and application features. Everyone who uses a stored process server or pooled workspace server has identical host-layer access to any SAS data that server retrieves. In a migrated deployment, access to most resources and data mirrors access in the original environment. To manage access to objects such as reports, stored processes, information maps, and data definitions, create custom folders that reflect
the distinctions that you want to make. See Chapter 6, “Access to Metadata Folders,” on page 69 .
9. To fully protect SAS data sets, you must address host access. See
Chapter 8, “Access to SAS Data,” on page 89 .
See Also
“Permissions by Object Type” on page 48
41
Part 2
Authorization
Chapter 5
Metadata Authorization Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Chapter 6
Access to Metadata Folders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Chapter 7
Access to Server Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Chapter 8
Access to SAS Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Chapter 9
Batch Tools for Metadata Authorization . . . . . . . . . . . . . . . . . . . . . . . . .
42
43
Chapter 5
Metadata Authorization Model
Permission Tips for Selected Content and Data Objects . . . . . . . . . . . . . . . . . . . . . 49
Permission Tips for Selected System and Administrative Objects . . . . . . . . . . . . . . 52
Working with SAS OLAP Shared Dimensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
44
Chapter 5 • Metadata Authorization Model
Authorization Overview
The platform provides a metadata-based authorization layer that supplements protections from the host environment and other systems. You can use the metadata authorization layer to manage access to almost any metadata object (for example, reports, data definitions, information maps, jobs, stored processes, and server definitions).
Across authorization layers, protections are cumulative. In order to perform a task, a user must have sufficient access in all applicable layers.
In the metadata layer, the following permissions are always enforced:
• the ReadMetadata permission (RM), which controls the ability to see an object
• the WriteMetadata permission (WM), which controls the ability to update or delete an object
Other permissions are specialized and affect only certain types of objects.
CAUTION:
In the metadata authorization layer, not all permissions are enforced for all items.
It is essential to understand which actions are controlled by each permission.
See “Use and Enforcement of Each Permission” on page 67 .
CAUTION:
Some clients enable power users to create and run SAS programs that access data directly, bypassing metadata-layer controls.
It is essential to manage
Three Levels of Granularity
You can set permissions at the following levels of granularity:
• Repository-level controls function as a gateway and as a parent-of-last-resort.
Repository-level controls are managed from the permission pattern of the repository
ACT (Default ACT). All registered users should have ReadMetadata and
WriteMetadata permissions in the foundation repository ACT’s permission pattern.
• Object-level controls manage access to a specific object such as a report, an information map, a stored process, a table, a column, a cube, or a folder. You can define resource-level controls individually (as explicit settings) or in patterns (by applying access control templates).
• Fine-grained controls affect access to subsets of data within a resource. To establish fine-grained controls, you add constraints called permission conditions to explicit grants of the Read permission.
Two Relationship Networks
Permission settings are conveyed across two distinct relationship networks:
Two Relationship Networks
45
• In object inheritance, permissions that you set on one object can affect many other objects. For example, a report inherits permissions from the folder in which the report is located. This network is a simple folder tree, with exceptions such as the following:
• The root folder is not the ultimate parent. This folder inherits from the repository
(through the permission pattern of the repository ACT).
• The root folder is not a universal parent. Some system resources (such as application servers, identities, and ACTs) are not in the folder tree. For these items, the repository ACT is the immediate and only parent.
• Inheritance within a table or cube follows the data structure. For example, neither table columns nor cube dimensions have folders as immediate parents. Instead, a column inherits from its parent table and a dimension inherits from its parent cube.
• Inheritance does not flow through specialty folders such as favorites folders, virtual folders, or search folders.
• In the identity relationships network, permissions that you assign to one identity can affect many other identities. For example, if you grant a group access to a report, that grant applies to everyone who is a member of the group. This relationship network is governed by a precedence order that starts with a primary (usually individual) identity, can incorporate multiple levels of nested group memberships, and ends with implicit memberships in SASUSERS and then PUBLIC.
The following figures depict the relative priority and specificity of access controls within each of these networks. From top to bottom, the elements in each figure are ordered as follows:
• from highest precedence (hardest to override) to lowest precedence (easiest to override)
• from narrowest impact (most specific) to broadest impact (least specific)
Figure 5.1 Priority and Specificity in Object Inheritance higher precedence narrower impact direct controls
Explicit Controls
(ACEs)
Access Control Templates
(ACTs)
Controls on Parent Objects
Repository-Level Controls (Default ACT)
lower precedence broader impact
46
Chapter 5 • Metadata Authorization Model
Figure 5.2 Priority and Specificity in Identity Hierarchy higher precedence
User
Direct Memberships
Indirect Memberships
implicit memberships
SASUSERS (registered users)
PUBLIC (everyone that can log on)
lower precedence narrower impact broader impact
Object Inheritance
In the metadata layer, parent objects convey their effective permissions to child objects.
Children inherit the net effect of their parents' access controls, not the access controls themselves. The following figures depict inheritance paths in a foundation repository.
The arrows in the first figure flow from child to parent (for example, a table inherits effective permissions from its parent folder). The arrows in the second figure flow from parent to child (for example, a folder conveys its effective permissions to the items that it contains).
Figure 5.3 Inheritance Paths (Separated View)
Syste m Objects
Repository
ACT
BI Conten t Relational Dat a OLAP Dat a
User
Group
Role
ACT
Application
Server
Root
Folder
Logical
Server
Parent
Folder
Server
Connection
Stored
Process
Report
Information
Map
Table Library
Column
Shared
Dimension
Level
OLAP
Schema Cube
Hierarchy
Measure
Dimension
Level
Hierarchy
Object Inheritance
47
Figure 5.4 Inheritance Paths (Integrated View)
Repository ACT
ACT
User, Group, Role
Application
Server
Root Folder
Logical
Server
Server
Parent Folder
Stored Process
Report
Information Map
Table
Library
Column
Shared Dimension
Level
Hierarchy
OLAP Schema
Connection
Cube
Measure
Dimension
Level
Hierarchy
Here are some details about the preceding figures:
• The depicted folder structure is arbitrary and intended only to show the security relationships between different types of objects.
• Not all object types are depicted.
T I P
In SAS Management Console, you can trace an object’s inheritance by clicking Advanced on the object's Authorization tab. This feature is available to only unrestricted users.
• The root folder represents the top of the folder tree for the foundation repository (the
SAS Folders node).
• The root folder inherits settings from the permission pattern of the repository ACT
(which is usually named Default ACT).
48
Chapter 5 • Metadata Authorization Model
• Any custom repositories are represented as folders (immediate children of the foundation root folder). Although these folders inherit permissions from both the foundation root folder and the repository ACT of the custom repository, access to objects within the custom repository branch should be managed from the folder side whenever possible.
• In some clients, your My Folder is displayed directly below the root folder. This is just a shortcut for accessing your personal content area. This folder is not an immediate child of the root folder.
• In general, specialized folders (such as search folders, favorites folders, and virtual folders) do not convey permissions to the objects that they contain. An exception is that a favorites folder does convey permissions to any child favorites folders
(favorites groups) that it contains.
• The figures show users, groups, and roles inheriting repository-level permissions. In some clients, the authorization information for a user, group, or role reflects special rules that protect identity definitions.
• The inheritance path for metadata-bound data is different. Secured table objects inherit from secured library objects, not from a parent folder. See “Object Creation,
Location, and Inheritance” in SAS Guide to Metadata-Bound Libraries .
See Also
•
“Identity Hierarchy” on page 17
•
“Authorization Decisions” on page 60
Permissions by Object Type
Introduction
The following figure depicts common objects and uses arrows to indicate objects whose permission settings are most likely to need an adjustment. Here are some details about the figure:
• Most of the up arrows indicate a need to extend the ability to read data. In the interest of a more secure initial deployment, Read access to data is not granted by default.
• Each down arrow indicates a need to limit the ability to modify or delete a system object. Other system items are protected by default.
• MLE refers to the metadata LIBNAME engine. The MLE items are listed to highlight the elevated permission requirements that apply when data is accessed through this engine.
• The intent of the figure is to highlight the adjustments that are most commonly needed. In general, it is preferable to set permissions on a parent (such as a folder) rather than on each individual object (such as a report).
Permissions by Object Type
49
Figure 5.5 Common Access Adjustments by Object Type
Le gend:
Increase access to make fully available.
BI Conte nt
Folder
Information map
Report
Stored process
Publication channel
Dashboard
Reduce access to improve security.
D ata
Folder
MLE library
Other library
MLE table
Other table
OLAP schema
Cube
S yste m
Repository
AC T s
Repository
Predefined
User-defined
Serve rs
Application
Logical
Server
Identi ties
User
Group
Role
The following tables provide object-specific instructions and tips. The purpose of the tables is to highlight special requirements for common items.
T I P
For any change-managed areas or resources, change-managed users should have
CheckInMetadata (CM) permission (instead of WM or WMM). See “Setting Up
Change Management” in SAS Intelligence Platform: Desktop Application
Administration Guide
.
Permission Tips for Selected Content and Data Objects
Table 5.1 Permissions Tips: Selected Content and Data Objects
Root folder
The root folder (the SAS Folders node) is an important point of control for narrowing WM. Unlike other folders, the root folder does not have the WriteMemberMetadata permission in its permission list. This is because the root folder is a software component object, not a true folder. The root folder can contain only other folders.
50
Chapter 5 • Metadata Authorization Model
Folder
For folders, follow these guidelines:
• Provide users with a clear path of grants of ReadMetadata permission to all of the content that they access. This is a navigational requirement. To browse past a folder, you need ReadMetadata permission for that folder.
• To enable users to contribute objects to a folder, grant them WriteMemberMetadata permission on that folder.
• On a folder, grant WriteMetadata permission only to users who should be able to delete, move, or rename that folder.
• Do not assume that every permission that is listed for a folder is relevant for every object in that folder.
• If you deny ReadMetadata permission on a folder, make sure that you do not prevent the SAS Trusted
User from having that permission on all cubes and schemas within that folder. One approach is to give the SAS System Services group a grant of ReadMetadata permission on the folder. This preserves necessary access for this privileged service identity.
See Chapter 6, “Access to Metadata Folders,” on page 69
.
Information map
To access any data through an information map, you need Read permission on that information map. You can manage Read permission globally or selectively. For example:
• A broad approach is to grant Read permission to SASUSERS in the repository ACT's permission pattern, or on the top folder.
• A narrow approach is to grant Read permission to smaller groups on specific subfolders or even specific information maps.
Report
Stored process
Channel
If a user cannot run a report, check these things:
• Does the user have Read permission for the underlying information map?
• Does the user have Read permission for any underlying OLAP cube?
• Does the user have Read permission for any underlying relational data that is accessed via the MLE?
• Does the SAS Trusted User have ReadMetadata permission for any underlying cube?
• Does the account that retrieves any underlying SAS data have physical access to that data?
It is especially important to manage ReadMetadata permission to pregenerated reports, because those reports can contain embedded data. Any user who views a pregenerated report sees the same data, regardless of his or her permissions to the underlying tables or cubes.
To run a stored process, you need ReadMetadata permission on the stored process. If you can see a stored process but cannot run it, you might lack the necessary grant of Read permission for the underlying data.
To register a stored process, you need WriteMetadata permission for the target application server.
The method that you use to make a stored process available can affect data retrieval and security. For example, in the standard configuration, a stored process that is assigned to a workspace server and embedded in an information map retrieves SAS data under the pooled workspace server's host identity.
However, if a user opens that same stored process directly (for example, as a report in SAS Web Report
Studio), the host identity of the requesting user (or group) retrieves the data.
In a new deployment, only the SAS Administrators group can add channels, subscribers, and content. To enable all registered users to publish content to a particular channel, navigate on the Folders tab to
System ð Publishing ð Channels and grant Write and WriteMetadata permissions to SASUSERS on that channel (WM is required only if a channel has an archive persistent store).
To enable all registered users to add channels or subscribers, grant WriteMemberMetadata permission on the relevant parent folder (for example, on the System ð Publishing ð Subscribers ð Content
Subscribers folder).
Permissions by Object Type
51
See “Implementing Security for SAS BI Dashboard” in SAS Intelligence Platform: Web Application
Administration Guide
.
Dashboard
Schema
Cube
Cube component
To associate an OLAP schema with an application server, you need WriteMetadata permission for the schema and the server. To add cubes to a schema, you need WriteMetadata permission for the schema and
WriteMemberMetadata permission for the target folder.
The SAS Trusted User must have ReadMetadata permission for all OLAP schemas and cubes. This access is usually granted through the SAS System Services group.
To register cubes, you need WriteMetadata permission for the OLAP schema and WriteMemberMetadata permission for the target folder. To associate a cube with a schema, you need WriteMetadata permission for the cube and the schema.
Read permission is enforced for cubes. Grant Read permission broadly (as described for information maps), or add narrower grants of Read permission on folders or individual cubes.
The SAS Trusted User must have ReadMetadata permission for all OLAP schemas and cubes. This access is granted through the SAS System Services group.
For cube components, Read permission is enforced. If you lack access to a measure that participates in a calculated measure, you can get unintended results.
There are also navigational requirements for Read permission on cube components. If a user does not have Read permission to a hierarchy, the user cannot navigate to the top levels within the hierarchy. If a user does not have Read permission to a particular level in a hierarchy, the user cannot navigate to the next level.
To set permissions on a shared dimension, navigate to it directly under its parent folder. You cannot set permissions on a shared dimension that you accessed by navigating within a cube.
You cannot delete a shared dimension that is in use (that is included in one or more cubes).
Shared dimension
To associate a library with an application server, you need WriteMetadata permission for the server (but not for the library). For a library that is accessed via the metadata LIBNAME engine (MLE), you need the
Create permission in order to add tables and the Delete permission in order to delete tables.
Library
Table
To associate a table with a library, you need WriteMetadata permission for the table and the library. For a table that is accessed via the MLE, you need Read permission in order to access data. The Create, Delete, and Write permissions affect your ability to add, update, or delete data. You can grant these permissions broadly (as described in the information map row) or precisely (for example, to a small group of users on a particular table).
The MLE does not support column-level access distinctions for the Read permission. Column-level access distinctions for the ReadMetadata permission are supported.
Column
52
Chapter 5 • Metadata Authorization Model
Permission Tips for Selected System and Administrative Objects
Table 5.2 Permissions Tips: Selected System and Administrative Objects
Repository
Application server
To monitor or operate servers other than the metadata server, you need the Administer permission on the server. (The metadata server requires the Metadata Server: Operation role instead of the Administer permission.)
To associate a stored process, OLAP schema, or library with an application server, you need
WriteMetadata permission for that application server. Certain service identities need ReadMetadata
permission to all server definitions. See Chapter 7, “Access to Server Definitions,” on page 81 .
To use a logical server, you need ReadMetadata permission for at least one of that server's connections.
This is called server access security. Certain service identities need ReadMetadata permission to logical server definitions. See
“Hide Server Definitions” on page 84 .
Logical server
On a foundation repository, all participating users should have repository-level ReadMetadata and
WriteMetadata permissions. Other predefined entries in the permission pattern of the repository ACT
(Default ACT) provide necessary administrative and service access.
Identity
User administration capabilities (from the Metadata Server: User Administration role) enable you to create, update, and delete users, groups, and roles. You can delegate management of an identity to someone who does not have user administration capabilities by adding explicit or ACT grants of
WriteMetadata permission in the identity's authorization properties. An identity's authorization properties have no effect on what that identity can do. You need ManageMemberMetadata permission to change the membership of the UserGroup and Role. ManageCredentialsMetadata enables you to manage accounts and trusted logins of User and UserGroup.
To create an ACT, you need repository-level WriteMetadata permission. Each predefined ACT is protected by direct access controls. ACTs that you create are not automatically protected. It is essential to add protections (direct controls in the ACT’s authorization properties) to any ACTs that you create.
ACT
Secured data folder
A secured data folder is a specialized container for secured library objects. In each metadata repository, the first secured data folder is Secured Libraries, within the System folder. See
Secured library object
A secured library object is the target to which a physical SAS library is bound. The Select permission on each secured library object manages Read access to its corresponding physical library. Secured library objects are stored beneath a repository’s Secured Libraries folder. See
“Metadata-Bound Libraries” on page 94 .
Secured table object
A secured table object is the target to which a physical SAS data set (a table or view) is bound. The Select permission on each secured table object manages Read access to its corresponding physical data set. Each secured table object is stored beneath a secured library object. See
“Metadata-Bound Libraries” on page
Permissions by Task
53
See Also
•
“Permissions by Task” on page 53
•
“Use and Enforcement of Each Permission” on page 67
Permissions by Task
Introduction
The following tables show required metadata layer permissions for selected tasks. For each task, a user must have the specified access to the specified metadata objects.
T I P
For any change-managed areas or resources, change-managed users should have
CheckInMetadata (CM) permission (instead of WM or WMM). See “Setting Up
Change Management” in SAS Intelligence Platform: Desktop Application
Administration Guide
.
Working with Folders
Table 5.3 Working with Folders
Task Repository Parent Folder
Add a folder
Delete a folder
RM, WM
RM
RM, WMM
*
RM, WMM
*
Rename a folder
Set folder permissions
Add an item to a folder
Delete an item from a folder
RM
RM
RM, WM
RM
Copy/export items
Paste/import items
RM
RM, WM
*
If the parent folder is the root folder , you need RM, WM on the root folder.
RM
RM
RM
RM
RM
RM
Folder
-
RM, WM
RM, WM
RM, WM
RM, WMM
RM, WMM
RM
RM, WMM
Item
-
-
-
-
-
RM, WM
RM
-
54
Chapter 5 • Metadata Authorization Model
Working with Reports
Table 5.4 Working with Reports
Task Repository
Parent
Folder Report
Stored
Process
*
Information
Map
*
Data
RM, R
**
Create and save a new report
Delete a report
View or refresh a report
RM, WM
RM
RM
RM,
WMM
RM,
WMM
RM
-
RM, WM
RM
RM
-
RM
RM, R
-
RM, R
View a batch report
Edit or rename a report
RM
RM
RM
RM
RM
RM, WM
-
-
-
-
Set report permissions RM RM RM, WM -
*
**
This is not a required element for a report.
The Read permission is required for data that is accessed through the metadata LIBNAME engine or the OLAP server.
-
RM, R
-
-
-
**
Working with Information Maps
Table 5.5 Working with Information Maps
Task Repository
Parent
Folder
Information
Map
Stored
Process
*
Create and save a new information map
Delete an information map
Set information map permissions
RM, WM
RM
RM
RM, WMM
RM, WMM
RM
-
RM, WM
RM, WM
RM
-
-
Edit or rename an information map
Run queries in an information map
RM
RM
RM
RM
RM, WM
RM, R
-
RM
*
**
This is not a required element for an information map.
The Read permission is required for data that is accessed through the metadata LIBNAME engine or the OLAP server.
Data
RM, R
**
-
-
-
RM, R
**
Permissions by Task
55
Working with Stored Processes
Table 5.6 Working with Stored Processes
Task Repository
Parent
Folder
Application
Server
Stored
Process Data
Register a stored process RM, WM RM, WMM
RM, WM
*
-
Delete a stored process
Set stored process permissions
RM
RM
RM, WMM
RM
RM, WM
*
RM
RM, WM
RM, WM
-
-
Run a stored process RM RM RM RM
*
**
The WM permission is required for stored processes with physical source code files and only on the SAS Application Server.
The Read permission is required for data that is accessed through the metadata LIBNAME engine or the OLAP server.
RM, R
**
Working with Publishing Channels
Table 5.7 Working with Publishing Channels
Task Repository Parent Folder
Add a channel or subscriber
Delete a channel or subscriber
RM, WM
RM
Edit a channel or subscriber
Publish content to a channel
RM
RM, WM
*
*
**
WM is required if the channel has an archive persistent store.
Content is published to only those subscribers for whom you have RM.
RM, WMM
RM, WMM
RM
RM
Channel
-
RM, WM
RM, WM
RM, W, WM
*
Subscriber
-
RM, WM
RM, WM
RM
**
Working with Tables
Table 5.8 Working with Tables
Task
Register a table
Delete a table
Set table permissions
Access table data
Repository
RM, WM
RM
RM
RM
Server
*
RM
RM
-
RM
Library
RM, WM
RM, WM
RM
RM
Parent
Folder
RM, WMM
RM, WMM
RM
RM
Table
-
RM, WM
RM, WM
RM, R
**
Column
-
-
-
RM
56
Chapter 5 • Metadata Authorization Model
Task Repository Server
*
Library
Parent
Folder
Register a library RM, WM RM, WM RM, WMM
*
**
SAS Application Server
The Read permission is required for data that is accessed through the metadata LIBNAME engine.
Table
-
Column
-
Note: The preceding information applies to working with traditional library and table
objects. If you are using metadata-bound data (secured library objects and secured table objects), see “Permissions for Metadata-Bound Data” in SAS Guide to
Metadata-Bound Libraries
instead.
Working with SAS OLAP Cubes
Table 5.9 Working with SAS OLAP Cubes
Task Repository Server
*
Schema
Parent
Folder Cube
Register a cube
Delete a cube
Rebuild a cube
Refresh a cube
RM, WM
RM
RM
RM
RM
RM
RM
RM
RM, WM
RM, WM
RM
RM
RM, WMM
Access cube data
Register a schema
Set cube permissions
RM
RM, WM
RM
RM
RM, WM
-
RM
-
RM
RM
RM, WMM
RM
Use the OLAP Server Monitor RM A -
*
**
SAS Application Server
The Read permission is required for data that is accessed through the metadata LIBNAME engine.
RM, R
-
RM, WM
-
-
RM, WMM RM, WM
RM RM, WM
RM RM, R
Source
Data
RM, R
**
-
RM, R
**
RM, R
**
-
-
-
-
Working with SAS OLAP Shared Dimensions
Table 5.10 Working with SAS OLAP Shared Dimensions
Task
Register a shared dimension
Delete a shared dimension
***
Repository
RM, WM
RM
Server
*
RM
RM
Schema
RM, WM
Parent
Folder
RM, WMM
Cube
-
Shared
Dimension
-
RM, WM RM, WMM RM, WM
Source
Data
RM, R
**
-
Permission Origins
57
Task Repository Server
*
Schema
Parent
Folder Cube
Shared
Dimension
Rebuild a shared dimension
RM RM RM RM -
Use a shared dimension
RM RM RM RM RM, R
*
**
***
SAS Application Server
The Read permission is required for data that is accessed through the metadata LIBNAME engine.
You cannot delete a shared dimension that has any associations to any cubes.
RM, WM
RM
Source
Data
RM, R
**
-
See Also
•
“Permissions by Object Type” on page 48
•
“Use and Enforcement of Each Permission” on page 67
Permission Origins
Introduction to Permission Origins
The permission origins feature identifies the source of each effective permission.
Permission origins answers the question: Why is this identity granted (or denied) this permission?
In the origins answer, only the controlling (winning, highest precedence) access control is shown. If there are multiple tied winning controls, they are all shown. Other, lower precedence controls are not shown in the origins answer.
Simple Permission Origins
The following table provides simple examples of permission origins answers. In each example, we are interested in why UserA has an effective grant on FolderA. In each example, UserA is a direct member of both GroupA and GroupB. Each row in the table is for a different (independent) permissions scenario. In the table, the first column depicts the contents of the Origins window. The second column interprets the information.
Table 5.11 Origins: Simple Examples
Origins Information
UserA [Explicit]
Source of UserA's Effective Grant on FolderA
On FolderA, an explicit grant for UserA
GroupA [Explicit]
On FolderA, an explicit grant for GroupA
58
Chapter 5 • Metadata Authorization Model
Origins Information
GroupA [Explicit]
GroupB [Explicit]
GroupA [ACT: GroupARead]
SASUSERS [ACT: GenRead]
GroupA [ACT: GroupARead]
GroupB [ACT: GroupBRead]
GroupA [ACT: GroupABRead]
GroupB [ACT: GroupABRead]
UserA is unrestricted.
This setting mirrors the WriteMetadata setting.
Source of UserA's Effective Grant on FolderA
On FolderA, explicit grants for GroupA and GroupB
Note: Two settings are shown because they are tied and they both win (UserA is a direct member of GroupA and GroupB).
On FolderA, an ACT pattern grant for GroupA (from a directly applied ACT)
On FolderA, an ACT pattern grant for SASUSERS (from a directly applied ACT)
On FolderA, ACT pattern grants for GroupA and GroupB
(from two different directly applied ACTs).
Note: Two settings are shown because they are tied and they both win (UserA is a direct member of GroupA and GroupB).
On FolderA, ACT pattern grants for GroupA and GroupB
(from the same directly applied ACT).
Note: Two settings are shown because they are tied and they both win (UserA is a direct member of GroupA and GroupB).
UserA’s status as an unrestricted user (someone who is unrestricted is always granted all permissions)
The WriteMetadata setting on FolderA. To investigate further, examine the origins for that setting.
Inherited Permission Origins
In many cases, the controlling setting is not on the current object. Instead, the controlling setting is defined on a parent object and inherited by the current object.
The following table provides examples in which the controlling setting comes from a parent object. Because the source of the effective permission is a parent object, the answer must identify which parent object has the controlling setting. For this reason, the origins answers in the following examples identify both a particular parent object (the object that has the controlling setting) and the controlling setting itself.
In each example, we are interested in why UserA has an effective grant on FolderA. In each example, UserA is a direct member of both GroupA and GroupB. Each row in the table is for a different (independent) permissions scenario. In the table, the first column depicts the contents of the Origins window. The second column interprets the information.
Table 5.12 Origins: Inheritance Examples
Origins Information Source of UserA's Effective Grant on FolderA
On ParentFolderA, an explicit grant for UserA
ParentFolderA
UserA [Explicit]
Types of Access Controls
59
Origins Information
ParentFolderA
GroupA [Explicit]
Source of UserA's Effective Grant on FolderA
On ParentFolderA, an explicit grant for GroupA
ParentFolderA
GroupA [Explicit]
GroupB [Explicit]
ParentFolderA
GroupA [ACT: GroupARead]
On ParentFolderA, explicit grants for GroupA and GroupB
On ParentFolderA, an ACT pattern grant for GroupA (from a directly applied ACT)
GreatGrandParentFolderA
SASUSERS [ACT: GenRead]
ParentFolderA
GroupA [ACT: GroupARead]
GroupB [ACT: GroupBRead]
On GreatGrandParentFolderA, an ACT pattern grant for SASUSERS
(from a directly applied ACT)
On ParentFolderA, ACT pattern grants for GroupA and GroupB (from two different directly applied ACTs)
GrandParentFolderA
GroupA [ACT: GroupABRead]
GroupB [ACT: GroupABRead]
On GrandParentFolderA, ACT pattern grants for GroupA and GroupB
(from the same directly applied ACT).
SAS Folders
SASUSERS [Explicit]
Default ACT [CustomRepositoryA]
SASUSERS [Pattern]
On the SAS Folders node, an explicit grant for SASUSERS.
Also, in CustomRepositoryA’s default ACT, a pattern grant for UserA.
Note: In this example, FolderA is within a custom repository, so it inherits from both the SAS Folders node and the custom repository’s default ACT pattern. Two settings are shown because they are tied and they both win.
Types of Access Controls
explicit controls
A user has an explicit control on an object if the user is directly and individually granted or denied a permission on the object. Explicit settings have the highest precedence. However, managing a large number of explicit controls for individual users can be cumbersome. For greater efficiency, we recommend that you set explicit controls for groups, use ACTs, and rely on inheritance.
60
Chapter 5 • Metadata Authorization Model
ACT settings
A user has an ACT setting on an object if an ACT that is applied to the object has a permission pattern that explicitly grants or denies the relevant permission to the user.
Each ACT adds its pattern of grants and denials to the settings for each object to which the ACT is applied.
indirect settings
One way that a user can have an indirect setting on an object is if the user belongs to a group that has an explicit or ACT setting on the object. Another way that a user can have an indirect setting on an object is through access control inheritance. Inherited settings come from a parent object (such as a folder). Inherited settings matter only if there are no relevant direct controls on the target object. The term “indirect settings” is also used to refer to a WriteMemberMetadata setting that mirrors the
WriteMetadata setting, and to grants that come from a special status (such as being unrestricted).
permission conditions
Permission conditions constrain explicit grants of the Read permission, so that only a subset of the target data is available. For example, you can set permission conditions on OLAP dimensions (limiting access to members) and information maps (limiting access to rows).
Authorization Decisions
Access Control Evaluation Process
The relative precedence of each access control is based on where it set, who it is assigned to, and how it is set. The following list summarizes how the metadata server evaluates all relevant access controls to reach an authorization decision:
1. Direct controls (permissions that are set directly on the target object) are examined.
• Any conflicts that arise from group membership are resolved by the identity hierarchy. For example, an explicit control that is assigned to a user overrides a conflicting explicit control that is assigned to a group to which the user belongs.
• If there is a conflict between an explicit control and an ACT setting at the same level in the identity hierarchy, then the explicit control takes precedence.
• If there is a conflict between two explicit controls (or two ACT settings) at the same level in the identity precedence hierarchy, then the outcome is a denial.
• If one or more permission conditions have been defined, then the condition that is assigned at the highest level of identity precedence is applied. Other conditions that also apply to a user because of group memberships do not provide additional, cumulative access (unless there are multiple tied groups at the highest level of identity precedence).
• If there are no relevant direct controls, then the evaluation process continues.
2. The step 1 evaluation process is applied to the immediate parent of the target object.
For example, the immediate parent of a report is its folder. If no direct controls are found, each successive parent object is examined in turn.
Note: In the unusual circumstance in which an object has more than one immediate
parent, a grant from any inheritance path is sufficient to provide access.
Authorization Decisions
61
3. If no direct controls are found on the object or on any of its parent objects, the permission pattern of the repository ACT (Default ACT) is examined. The repository
ACT serves as the inheritance parent of last resort.
• If the repository ACT grants or denies the requested permission, then that grant or denial is determinative.
• If the repository ACT neither grants nor denies the permission, then the permission is denied.
• If there is no repository ACT, then the permission is granted. You should always have a designated repository ACT.
Precedence Principles and Examples
The following table summarizes the principles of the authorization decision process for a user (Joe) who is looking for an object (LibraryA or ObjectA):
Table 5.13 Precedence Principles for Authorization
Principle Example
*
Outcome
Settings on an object have priority over settings on the object's parents.
LibraryA has an explicit denial for PUBLIC. LibraryA's parent folder has an explicit grant for Joe.
Joe cannot see
LibraryA.
Conflicting settings on an object are resolved by identity precedence.
LibraryA has an explicit denial for GroupA and an explicit grant for GroupAA. Joe is a direct member of GroupA.
GroupA is a direct member of GroupAA.
Joe cannot see
LibraryA.
In an identity precedence tie, explicit settings have priority over ACT settings.
LibraryA has an ACT denial for GroupA and an explicit grant for GroupB. Joe is a direct member of both GroupA and
GroupB.
Joe can see
LibraryA.
In an identity precedence tie that is not resolved by the preceding row, the outcome is a denial.
A grant from any inheritance path can provide access.
LibraryA has an explicit denial for GroupA and an explicit grant for GroupB. Joe is a direct member of both GroupA and
GroupB.
Joe cannot see
LibraryA.
ObjectA has no explicit or ACT settings, one immediate parent that conveys a grant, and another immediate parent that conveys a denial.
**
Joe can see
ObjectA.
*
**
The settings described in this column are the only explicit or ACT settings for ReadMetadata permission on LibraryA (or ObjectA).
Having more than one immediate parent is not a common circumstance.
See Also
•
“Object Inheritance” on page 46
•
“Identity Hierarchy” on page 17
•
62
Chapter 5 • Metadata Authorization Model
Fine-Grained Controls for Data
What Are Fine-Grained Controls?
Business requirements often specify that different users should see different portions, or slices, of data. In some cases, the requirement is driven by the sensitive nature of data.
For example, company policy might state that each salesperson should be able to access only his or her own salary information. In other cases, the requirement is intended to prevent information overload. For example, each regional sales team within a national organization might be interested in only the sales trend information for their region.
Fine-grained access distinctions are frequently based on each user's place in an organizational structure such as a management hierarchy or a product matrix. The visibility of data can depend on a simple, site-specific condition such as a user's security clearance level, or on a more complex condition that consists of multiple filters.
You use fine-grained controls to specify who can access particular rows within a table or particular members within a cube dimension. These controls often subset data by a user characteristic such as employee ID or organizational unit. For example, a table that contains patient medical information might be protected by row-level permissions that enable each doctor to see only those rows that contain data about that doctor's patients.
Unlike other access controls, fine-grained controls are based on filters and rely on target data that is modeled to work with those filters. When fine-grained controls are used, there are three possible authorization decision outcomes for a user request to view data:
Grant
The requesting user can access all data.
Deny
The requesting user cannot access any data.
Conditional Grant
The requesting user can access only the data that meets specified filtering conditions.
What Implementations are Available?
Each of the following components offers an implementation of fine-grained controls:
BI row-level permissions provide filtering for SAS data sets and third-party relational data accessed through an information map. You define and assign the filters in SAS Information Map Studio or with the INFOMAPS procedure. See
SAS Guide to BI Row-Level Permissions
.
CAUTION:
Not all SAS clients require that users go through information maps in order to access data. Comprehensive security that incorporates BI row-level permissions requires a specialized configuration that is supported only by
SAS Web Report Studio.
Note: BI row-level permissions are primarily intended for use with SAS Web Report
Studio. For information about the row-level security feature that is available in
SAS Visual Analytics, see SAS Visual Analytics: Administration Guide.
Fine-Grained Controls for Data
63
SAS OLAP Server provides filtering for SAS OLAP data using MDX expressions. You define and assign the filters in SAS Management Console, SAS OLAP Cube Studio, or SAS
Data Integration Studio. See
SAS OLAP Server: User’s Guide
.
SAS Scalable Performance Data Server enables you to define database views that filter rows based on the user ID of the connecting client. This functionality is provided by the @SPDSUSR system variable.
The metadata-based identity-driven properties are not available in this implementation. See the SAS Scalable Performance Data Server: Administrator’s
Guide.
metadata-bound libraries provide two methods of filtering for metadata-bound tables. The first method is to provide users with views that include a portion of a table’s data without providing access to the underlying table. With this method, you can use the
_METADATA_AUTHENTICATED_USERID_ substitution parameter to filter rows based on the authenticated user ID of the requesting client. The second method, which is available with the first maintenance release for SAS 9.4, is to specify a filter in the secured table object’s authorization properties in SAS Management Console.
With this method, you can use the metadata-based identity-driven properties to filter rows. See SAS Guide to Metadata-Bound Libraries.
The following table compares the implementations:
Table 5.14 Comparison of Fine-Grained Control Implementations
Secure
Graphical Filter
Creation Authorization UI
Metadata-
Aware
Identity-
Driven Implementation
OLAP member-level
SPD Server row-level
BI row-level
Metadata-bound libraries
About Identity-Driven Properties
Overview of Identity-Driven Properties
It is often necessary to make per-person access distinctions for the rows in a table or the members in a dimension. You can make a separate filter for each user (such as
where name="joe"
). However, if you have more than a few users, this approach quickly becomes cumbersome. The more efficient alternative is to create a dynamic filter (such as
where name="&name;"
) that can discover and insert the correct, user-specific value into the WHERE expression each time access is requested.
To create a dynamic filter, use an identity-driven property as the value against which values in the target data are compared. This list explains how the substitution works:
1. Each identity-driven property corresponds to a characteristic (such as name, user ID, or external identity).
64
Chapter 5 • Metadata Authorization Model
2. Each user's values for these characteristics (such as
joe
,
WinXP\joe
, or
607189
) are stored in the metadata.
3. The identity-driven property is aware of the user ID with which a client authenticated and can locate information that is stored in the metadata for that user ID.
4. Each time it receives a request, the identity-driven property substitutes a userspecific value into the filter expression.
Note: This discussion is not applicable to the SPD Server, which has its own
implementation of identity-based filtering.
These are the most useful identity-driven properties:
SAS.Userid
returns an authenticated user ID, normalized to the uppercase format USERID or
USERID@DOMAIN.
SAS.ExternalIdentity
returns a site-specific value (for example, employee ID). This property is often useful because its values are likely to match user information in your data. An identity can have more than one external identity value. However, only the first value is returned. Unlike the values for other identity-driven properties, values for this
property are not always populated in the metadata. See “External Identities” on page
.
SAS.IdentityGroups
returns a list of the groups and roles that this identity belongs to (directly, indirectly, or implicitly). The list contains the group and role names, as displayed in the Name field on the General tab for each group or role.
SAS.PersonName
returns a user name, as displayed in the Name field in the user's general properties.
These identity-driven properties are also supported:
SAS.IdentityGroupName
returns a group name, as displayed in the Name field in the group's general properties. If a user logs on with an ID that is stored in a login on a group definition, then the name of the group that owns that login is returned. If a user logs on with a user ID that is not stored in the metadata, then the PUBLIC group is returned.
This property is useful only in the unusual circumstance where a user logs on with the user ID that is defined for a group login. In almost all cases, a user logs on with a user ID that is defined for an individual user definition. Not all applications allow a group to log on. This property is not supported if client-side pooling is used.
SAS.IdentityName
returns a user name or group name, as displayed in the Name field in the general properties for the user or group. This property is a generalization of
SAS.PersonName and SAS.IdentityGroupName.
Note: In certain circumstances, a connecting identity might not have a value for the
identity-driven property that you are using. This can happen with the
ExternalIdentity property (sometimes), the IdentityGroupName property (almost always), or the PersonName property (rarely). When a connecting user does not have a value for the property that a query uses, an empty string is returned or the query fails.
Fine-Grained Controls for Data
65
Syntax for Identity-Driven Properties
The syntax for using identity-driven properties varies depending on which implementation you are using:
• For information maps, use the following syntax:
&property-name
(for example,
&SAS.Userid
).
• For secured table objects in metadata-bound libraries, use the following syntax:
SUB::property-name
(for example,
SUB::SAS.Userid
).
Examples of Identity-Driven Properties
For example, to enable each user to see only his or her own salary information, you could give the PUBLIC group a filter that is based on the SAS.PersonName property. At run time, the SAS.PersonName value that is associated with the connected user ID is substituted into the filter. In this way, the query is modified as appropriate for each requesting client.
This table contains examples of filters that are based on identity properties, showing representations of both the generic form and how each filter would be modified when executed by a user named Harry Highpoint. Harry is a member of the ETL and
Executives groups. The example assumes that the customer has an employee information table named EmpInfo, which includes Name, Category, WinID, Department, and EmpID columns.
Table 5.15 Examples of Filters That Use Identity-Driven Properties
As Defined (Generic Form)
EmpInfo.WinID=&SAS.Userid
*
EmpInfo.WinID="SUB::SAS.Userid"
**
EmpInfo.EmpID=&SAS.ExternalIdentity
EmpInfo.EmpID="SUB::SAS.ExternalIdentity"
EmpInfo.Department IN &SAS.IdentityGroups
EmpInfo.Department IN ("SUB::SAS.IdentityGroups")
EmpInfo.Name=&SAS.IdentityName
EmpInfo.Name="SUB::SAS.IdentityName"
EmpInfo.Name=&SAS.PersonName
EmpInfo.Name="SUB::SAS.PersonName"
As Executed (Resolved Form)
Where EmpInfo.WinID="metadata_loginid"
Where EmpInfo.EmpID="metadata_externalid"
Where EmpInfo.Department IN ("metadata_group1"
"metadata_group2" "metadata_groupn")
Where EmpInfo.Name="metadata_identityname"
Where EmpInfo.Name="metadata_personname"
EmpInfo.Category=&SAS.IdentityGroupName
EmpInfo.Category="SUB::SAS.IdentityGroupName"
Where EmpInfo.Category="metadata_groupname"
***
*
**
***
Format to filter information maps.
Format to filter secured tables.
Because the user does not log on with a user ID that is stored as part of a group definition, the user has no value for this property. This either returns an empty string (in BI row-level permissions) or causes the query to fail (in other implementations).
66
Chapter 5 • Metadata Authorization Model
Batch Reporting Considerations
When you use fine-grained controls, it is essential to understand that only dynamically generated reports display data based on the access that is defined for the requesting user.
Static reports display data based on the access that is defined for the user ID that was used to generate the report. For example:
• Manually refreshed reports contain cached data (which can be updated by a user action in the report viewer).
• Pre-generated reports reflect the access of the user ID that was used to generate the report. Identity-specific access distinctions are preserved for pre-generated reports only if you define a separate report job for each user ID.
Permission Precedence Considerations
Note: This discussion applies to only OLAP member-level permissions and
authorization-based prefilters. It is not applicable to the SPD Server or to general prefilters in information maps.
Fine-grained controls are assigned to users or groups in the authorization properties of the target dimension, information map, or secured table object. These filters are available only to constrain either an explicit grant of the Read permission (on dimensions and information maps) or an explicit grant of the Select permission (on secured table objects). These filters are incorporated into the access control evaluation process as permission conditions.
A permission condition is applied only if it is on the setting that is closest to the requesting user. Other conditions that are relevant because of further-removed group memberships do not provide additional, cumulative access. If there is an identity precedence tie between multiple groups at the highest level of identity precedence, those tied conditions are combined in a Boolean OR expression. If the identity precedence tie includes an unconditional grant, access is not limited by any conditions. The following table provides examples:
Table 5.16 Precedence for Permission Conditions
Principle
If there are multiple permission conditions that apply to a user because of the user's group memberships, then the highest precedence identity controls the outcome.
Scenario Outcome and Explanation
A filter on InformationMapA limits
Read permission for GroupA.
Another filter on InformationMapA limits Read permission for the
SASUSERS group.
The user is a member of both GroupA and SASUSERS.
The user can see only the rows that
GroupA is permitted to see. GroupA has higher identity precedence than
SASUSERS, so the filters that are assigned to GroupA define the user's access.
If there are multiple permission conditions at the highest level of identity precedence, then any data that is allowed by any of the tied conditions is returned.
A filter on DimensionA limits Read permission for GroupA.
Another filter on DimensionA limits
Read permission for GroupB.
The user is a first-level member of both GroupA and GroupB.
The user can see any member that is permitted for either GroupA or GroupB.
Use and Enforcement of Each Permission
67
The following example describes the impact of identity precedence when a manager uses an information map that includes both of the following filters for a SALARY table:
• A permission condition that is assigned to the SASUSERS group gives each user access to his or her own salary information.
• A permission condition that is assigned to a Managers group enables each manager to see the salaries of the employees that he or she manages.
When the manager accesses the SALARY table, the filter that is assigned to the
Managers group is applied and the filter that is assigned to SASUSERS is ignored. This is because the manager's direct membership in the Managers group has higher identity precedence than the manager's implicit membership in the SASUSERS group. To avoid a situation in which managers can see their employees' salaries but each manager cannot see his or her own salary, you can use either of these approaches:
• Assign the filters to two groups that have the same identity precedence. For example, if you assign the first filter to a general purpose user-defined group (rather than to
SASUSERS), and you make each manager a direct member of that group, then managers have an identity precedence tie between that group and the Managers group. This situation causes the two filters to be combined for members of the
Managers group, enabling those users to see any row that is permitted by either filter.
• Define the Managers filter in a way that encompasses all of the rows that the managers should be able to see. In other words, combine (OR together) the
SASUSERS filter and the Managers filter.
Use and Enforcement of Each Permission
General-Purpose Permissions
The following table introduces the general-purpose permissions:
Table 5.17 General-Purpose Permissions
Permission
(Abbreviation)
ReadMetadata (RM)
Actions Affected
View an object. For example, to see a report, you need the ReadMetadata permission for that report.
WriteMetadata (WM) Edit, delete, or set permissions for an object. To delete an object, you also need the
WriteMemberMetadata permission for the object’s parent folder.
WriteMemberMetadata
(WMM)
Add an object to a folder or delete an object from a folder. To enable someone to interact with a folder's contents but with not the folder itself, grant WMM and deny WM.
*
CheckInMetadata (CM) Check in and check out objects in a change-managed area. The CheckInMetadata permission is applicable only in SAS Data Integration Studio.
*
A folder’s WMM settings mirror its WM settings unless the folder has a direct setting of WMM. A grant (or denial) of WMM on a folder becomes an inherited grant (or denial) of WM on the objects and subfolders in that folder. WMM is not inherited from one folder to another. WMM is not applicable to specialized folders (such as virtual folders).
68
Chapter 5 • Metadata Authorization Model
Specialized Permissions
The following table introduces some of the specialized permissions:
Table 5.18 Specialized Permissions
Permission
(Abbreviation)
Administer (A)
Create (C)
Read (R)
Write (W)
Delete (D)
ManageMemberMetadata (MMM)
ManageCredentialsMetadata (MCM)
Actions Affected
Operate (monitor, stop, pause, resume, refresh, or quiesce) certain SAS servers and spawners.
Add data through the metadata LIBNAME engine.
Read data through certain objects (for example, cubes, information maps, and tables that are accessed through the metadata LIBNAME engine).
Update data through certain objects (for example, data that is accessed through the metadata LIBNAME engine and publishing channels).
Delete data through the metadata LIBNAME engine.
Change the membership of the Group and Role. Cannot change security or other account attributes.
Manage accounts and trusted logins of User and Group. Cannot change security or other account attributes.
For more information, see the following documents:
•
SAS Language Interfaces to Metadata
(for information about the metadata
LIBNAME engine)
•
SAS Guide to Metadata-Bound Libraries
(for information about the Insert, Update,
Select, Create Table, Drop Table, and Alter Table permissions, and an additional use of the Delete permission)
69
Chapter 6
Access to Metadata Folders
Variation 1: Regional Separation, Designated Content Creators . . . . . . . . . . . . . .
Key Points about the Baseline ACT Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Further Considerations for Permissions on Folders . . . . . . . . . . . . . . . . . . . . . . . . .
Baseline ACTs
One approach to setting permissions on folders is to create a few general-use ACTs, and apply one or more of those ACTs to each folder that you need to secure. To grant access back to a particular group, supplement the ACT settings by adding explicit controls on the target folder. The examples in this chapter use three general-purpose ACTs. Each
ACT reduces a particular type of access down to a minimal level, so this chapter refers to these ACTs as baseline ACTs.
Hide gives SAS Administrators and service identities exclusive Read access to metadata
(limits visibility).
Table 6.1 Example: Pattern for the Hide ACT
Group
PUBLIC
SAS Administrators
Permission Pattern
Deny ReadMetadata
Grant
ReadMetadata
*
70
Chapter 6 • Access to Metadata Folders
Group
SAS System Services
Permission Pattern
Grant
ReadMetadata
**
*
**
This grant ensures that administrators can manage all metadata (for alternatives, see “Separated
).
This grant ensures that the SAS Trusted User (who is a member of SAS System Services) can read certain metadata on behalf of all users.
Protect gives SAS Administrators exclusive Write access to metadata (limits updates, deletions, and contributions).
Table 6.2 Example: Pattern for the Protect ACT
Group
PUBLIC
Permission Pattern
Deny WriteMetadata, WriteMemberMetadata,
CheckInMetadata, Write, Administer
SAS Administrators Grant WriteMetadata, WriteMemberMetadata,
CheckInMetadata, Write, Administer,
ReadMetadata
*
*
These grants ensure that administrators can manage all metadata (for alternatives, see “Separated
).
LimitData gives unrestricted users exclusive access to data (limits the availability of data that is accessed through information maps, the OLAP server, or the metadata LIBNAME engine).
Table 6.3 Example: Pattern for the LimitData ACT
Group
PUBLIC
Permission Pattern
*
Deny Read
*
This pattern is unusual in that it consists of a single setting. In the future, you might use this ACT to give a restricted user access to all data.
CAUTION:
Relational data that is accessed through other methods is unaffected by the
Read permission.
Do not rely exclusively on the metadata authorization layer to protect relational data. Use physical-layer protections (host or DBMS permissions) also.
See Also
“Use and Enforcement of Each Permission” on page 67
Example: Business Unit Separation
71
Example: Business Unit Separation
This example creates a secured custom branch with mutually exclusive access for two divisions. In this example, the project level folders are for only organizational purposes
(there are no access distinctions at the project level). The following figure depicts a partial group and folder structure.
Figure 6.1 Group and Folder Structure
The following table lists the protections for the first four folders in the branch:
Table 6.4 Mutually Exclusive Access
Folder
Direct Controls
ACTs Explicit Grants
DemoBranch
Protect
LimitData
DivisionA
Hide GroupA: +RM, +R
Project1
GroupA: +WMM
Project2
GroupA: +WMM
Here are some details about this example:
• ReadMetadata permission flows into the DemoBranch from its parent folder. In all of the examples in this chapter, the immediate parent to the DemoBranch is visible to all registered users (SASUSERS). This is a standard setting. Unless you apply the
Hide ACT to the top of your custom branch, the SASUSERS grant of ReadMetadata permission flows into your branch.
• The ability to create, manage, and delete content is shut off at the top (by the Protect
ACT on the DemoBranch). This constraint flows throughout the tree (except where you add supplemental grants of WriteMemberMetadata permission to specific functional groups on specific folders). Users do not add content or access data in the
DemoBranch, so no supplemental grants are needed on that folder.
72
Chapter 6 • Access to Metadata Folders
• Because you want to prevent members of GroupB from seeing the DivisionA branch, you apply the Hide ACT on that branch. You then add supplemental grants for
GroupA to restore their access.
• On DivisionA's project folders, you apply the Protect ACT to override GroupA's inherited grant of WriteMetadata permission (GroupA's grant of
WriteMemberMetadata permission on DivisionA becomes an inherited grant of
WriteMetadata permission on the immediate children of DivisionA). This prevents members of GroupA from renaming, deleting, or changing permissions on each project folder. You then add a supplemental grant of WriteMemberMetadata permission so that members of GroupA can contribute content in the project folders.
• On DivisionA's project folders, you do not apply the Hide ACT, because the
ReadMetadata access that flows in from the DivisionA folder is appropriate. In this example, the requirement is that all members of GroupA should be able to access content throughout the DivsionA branch.
• Notice that it is not necessary to use separate folders for each type of object.
• Any content contributors who register cubes must have WriteMetadata permission on the OLAP schema . By default, the schema is in the
SAS Folders/Shared
Data/SASApp - OLAP Schema
folder.
Variation 1: Regional Separation, Designated
Content Creators
This example eliminates the project folders and introduces the following requirements:
• Regional employees see only content for their region.
• A central group of managers see all content.
• A central group of content creators creates all content.
The following figure depicts a partial group and folder structure.
Figure 6.2 Group and Folder Structure
The following table lists the protections for the first four folders in the branch:
Variation 1: Regional Separation, Designated Content Creators
73
Table 6.5 Subgroup Separation and Limited Contribution
Folder
Direct Controls
ACTs Explicit Grants
DemoBranch
Protect
LimitData
DivisionA
Hide GroupA: +RM
ContentCreators: +RM
Managers:+RM
RegionA1
Hide
RegionA2
Hide
RegionA1: +RM, +R
ContentCreators: +RM, +R, +WMM
Managers:+RM, +R
RegionA2: +RM, +R
ContentCreators: +RM, +R, +WMM
Managers:+RM, +R
Notice that you are repeating many of the same explicit settings on each region, and that this will be the case throughout the DemoBranch. For greater efficiency and more centralized control, create a custom ACT (called RegionLevel) that provides the supplemental grants for your content creators (RM, R, WMM) and your managers (RM,
R). Remember to protect the ACT itself.
The following table lists the protections for the first four folders:
Table 6.6 Use a Supplemental ACT
Folder
Direct Controls
Baseline ACTs Supplemental Grants
DemoBranch
Protect
LimitData
DivisionA
Hide GroupA: +RM
ContentCreators: +RM
Managers:+RM
74
Chapter 6 • Access to Metadata Folders
Folder
RegionA1
RegionA2
Direct Controls
Baseline ACTs
Hide
Hide
Supplemental Grants
RegionA1: +RM, +R
RegionLevel
RegionA2: +RM, +R
RegionLevel
If you decide to offer content at the division level and you want that content to be available to only managers, you might make these changes:
• Create a DivisionLevel ACT with grants for Managers (RM, R) and ContentCreators
(RM, R, WMM). Apply that ACT to each division folder.
Note: This is the same pattern that you use for the RegionLevel ACT, so you could
instead simply use that ACT. In this example, you choose to create a separate
ACT because you anticipate that the requirements for division-level access and region-level access might diverge in the future.
• Apply the Protect ACT on each region folder (to take away the inherited grant of
WriteMetadata permission that content contributors inherit from their division-level grant of WriteMemberMetadata permission).
Note: If you choose to not do this, members of the content creators group can delete,
rename, or change permissions for the region folders.
The following table lists the protections for the first four folders in the branch:
Table 6.7 Accommodate Division-Level Content
Folder
Direct Controls
Baseline ACTs Supplemental Grants
DemoBranch
Protect
LimitData
DivisionA
Hide GroupA: +RM
DivisionLevel
RegionA1
Hide
Protect
RegionA1: +RM, +R
RegionLevel
RegionA2
Hide
Protect
RegionA2: +RM, +R
RegionLevel
Variation 2: Functional Separation
75
Variation 2: Functional Separation
This example includes three custom groups that represent users who have specific job responsibilities (Data Admins, Map Creators, and Report Creators). Each division folder has separate subfolders for different types of content (data definitions, information maps, report definitions, and stored processes). Supplemental Write access in each folder is limited to members of the appropriate functional groups as follows:
• Data administrators can work in each division's data definitions folder.
• Information map creators can work in each division's information maps folder.
• Report creators can work in each division's reports folder.
• Information map creators and report creators can work in each division's stored processes folder.
Note: Access for these functional groups is also limited by divisional affiliations. For
example, only DivisionA's report creators (and SAS Administrators) can add, update, and delete items in DivisionA's reports folder.
The following figure depicts a partial group and folder structure.
Figure 6.3 Group and Folder Structure
The following table lists the protections for the first six folders:
Table 6.8 Functional Separation
Folder
Direct Controls
ACTs Explicit Grants
DemoBranch
Protect
LimitData
DivisionA
Hide Managers: +RM, +R
GroupA: +RM, +R
76
Chapter 6 • Access to Metadata Folders
Folder
data definitions
information maps
reports
Direct Controls
ACTs
Protect
*
Protect
*
Protect
*
Explicit Grants
Data Admins: +WMM
Map Creators: +WMM
Report Creators: +WMM
stored processes
Protect
*
Map Creators: +WMM
Report Creators: +WMM
*
You do not strictly need the Protect ACT here, because protections flow through from the DemoBranch folder (and are not interrupted by any supplemental grants of WM or WMM on higher level folders). However, you do need the supplemental grants of WMM on these folders, so you might choose to also apply the Protect ACT here for clarity.
Here are some details about this example:
• On DivisionA, users who are not in the Managers group, GroupA, or SAS
Administrators are locked out. Such users cannot see or traverse DivisionA or any folders in the DivisionA branch. Managers and members of GroupA have supplemental grants of ReadMetadata and Read permissions on DivisionA; these grants flow through the entire DivisionA branch.
Note: In order to view a DivisionA report, you need ReadMetadata permission and
(in most cases) Read permission to the report's underlying components (such as information maps, stored processes, and data).
• On each content type folder, one or more functional groups has a supplemental grant of WriteMemberMetadata permission. However, functional group access is also constrained by division-level ReadMetadata access. For example, even though all report creators have WriteMemberMetadata permission on the DivisionA reports folder, only those report creators who are members of GroupA can see the folder.
• Nothing in these settings prevents creation of a certain type of item. For example, a map creator can create a report and add that report to the map folder or the stored processes folder. Here are some techniques for increasing control:
• You can use roles to limit the ability to create reports in SAS Web Report Studio.
• You should limit the ability to register libraries, stored processes, and OLAP schemas by managing WriteMetadata permission on application servers.
• You can use the map accessibility check configuration option to limit the locations from which SAS Web Report Studio will use relational maps.
• You might choose to put the supplemental grants of WriteMemberMetadata permission in additional ACTs (such as MapCreators ACT, ReportCreators ACT) instead of using explicit settings on each folder. Using ACTs is more centralized, which is beneficial in accommodating later changes to the pattern. Of course, the more divisions you have, the more work it would be to update explicit settings, so the more valuable it would be to centralize settings with ACTs.
• Consider how you would handle an exception requirement. For example, to let a particular data administrator who is not in GroupA contribute to the DivisionA data definitions folder, you might give that user a clear path of grants of ReadMetadata
Key Points about the Baseline ACT Approach
77
permission to the folder and supplemental grants of WriteMemberMetadata and Read permissions on the folder.
Key Points about the Baseline ACT Approach
• In general, it is not necessary to add protection to the predefined folders (such as the root folder, each user's My Folder, and the System folder). These folders are protected by default.
• The preceding examples do not add grants of ReadMetadata permission on the
DemoBranch folder because that folder's parent (the SAS Folders node) is visible to all registered users. The grant of ReadMetadata permission to SASUSERS on the
SAS Folders node flows into the DemoBranch.
• To hide a branch, apply the Hide ACT to the relevant folder and grant back
ReadMetadata permission to any groups who should have access. Remember that a denial of ReadMetadata permission on a folder prevents navigation to content and folders that are below the hidden folder (and that this cannot be mitigated by a grant of ReadMetadata permission on a lower-level folder or other item).
• To enable a group to contribute content to a particular folder, give that group a grant of WriteMemberMetadata permission on that folder and consider applying the
Protect ACT to each immediate subfolder.
Note: This protects lower-level folders that inherit the contributing group's
WriteMemberMetadata permission grant (from the parent folder) as a grant of
WriteMetadata permission (on the child folder). If you do not protect a child folder, the contributing group can rename, delete, or change permissions on that folder. To enable content contributions to the child folder, grant
WriteMemberMetadata permission on that folder (and consider repeating the denial of WriteMetadata permission on any subfolders that are below that folder).
Note: Any content contributors who register cubes must have WriteMetadata
permission on the OLAP schema. By default, the schema is in the
SAS
Folders/Shared Data/ SASApp - OLAP Schema
folder.
• If you want a group to access data through cubes, information maps, or the metadata
LIBNAME engine, give that group a grant of Read permission on the folder that contains the relevant items. For any folder where you add a grant of Read permission, determine whether you want that grant to flow through to the subfolders.
Apply the LimitData ACT to any immediate subfolders where you want to prevent data access. The navigational requirement for a clear path of ReadMetadata permission does not apply to the Read permission.
• The examples in this chapter ensure that content contributors have Read permission to the resources that they use. If you have a content contributor who is not also a content consumer, you can choose to not provide Read permission. For example, if you give a user ReadMetadata permission but not Read permission to an information map, that user can still design the map (but cannot perform tasks such as testing queries).
• You might be able to save some time by using this technique:
1. Create one sub-branch (for example, the DivisionA branch in the preceding examples). The folders should have appropriate permission settings but contain no content.
78
Chapter 6 • Access to Metadata Folders
2. Copy that empty sub-branch to create the other branches (for example, to create the DivisionB folders). Change the folder name on the copy and update the settings as necessary. Often, you only have to remove and add a few supplemental grants.
Note: If this technique proves useful, consider keeping an empty template branch
that is visible only to administrators. You can use the template if you need to add more branches later.
Further Considerations for Permissions on
Folders
Consolidation of ACTs
In general, consolidation (using one pattern in all of the places where it is appropriate) is beneficial, because it simplifies management. However, it might be appropriate to maintain two ACTs that have the similar patterns in circumstances such as these:
• You anticipate that access requirements might diverge. For example, if you think you will eventually separate folder administration from server administration, you might create a SystemProtect ACT for items that are not in the folder tree.
• You want to use a pattern that is similar to but not exactly the same as one of the predefined ACTs. For example, the baseline Hide ACT is not very different from the predefined Private User Folder ACT. We strongly recommend that you do not modify or delete the predefined ACTs, because these ACTs are an integral part of the protections that are set up for you during installation. The usage of each predefined
ACT requires certain settings. Modifying the settings on a predefined ACT can compromise the security that that ACT provides.
Note: The examples in this chapter do not demonstrate use of an ACT to protect other
ACTs. Consider returning to each ACT's Authorization tab, removing the explicit controls, and instead applying an ACT such as the Protect ACT.
Separated Administration
If you need to separate administration privileges by department, the approach in this chapter is not granular enough. If you do not want the SAS Administrators group to have universal access, consider creating parallel sets of baseline ACTs.
For example, to separate administration for an East region and a West region, you might create ACTs such as Hide_East, Hide_West. In each baseline ACT pattern, you would replace the SAS Administrators group with a narrower administrative group (for example, East_Admins, West_Admins). The denials to PUBLIC and grants to the SAS
System Services group would not change. Any unrestricted users can still access everything.
End Users, Folders, and Permissions
Proper use of the WriteMetadata and WriteMemberMetadata permissions protects a folder structure. Keep in mind that end users can affect access to content as follows:
Further Considerations for Permissions on Folders
79
• A user who can update an item can add settings on that item. You cannot prevent this by limiting the availability of SAS Management Console because users can set permissions in other applications (for example, SAS Information Map Studio, SAS
OLAP Cube Studio, SAS Data Integration Studio, SAS Enterprise Guide, and the
SAS Add-In for Microsoft Office).
• A user who can contribute items to a folder can also add subfolders below that folder. You cannot prevent this by limiting the availability of SAS Management
Console because users can add folders in other applications (for example, SAS
Information Map Studio, SAS OLAP Cube Studio, SAS Data Integration Studio,
SAS Enterprise Guide, and the SAS Add-In for Microsoft Office).
• If you give someone CheckInMetadata permission on a folder, that person can update or delete the folder (through change management activities), as well as check in content to that folder. Change management is an optional feature that is available only for SAS Data Integration Studio.
80
Chapter 6 • Access to Metadata Folders
81
Chapter 7
Access to Server Definitions
Protect Server Definitions
Introduction
In a new deployment, all registered users can update and delete server definitions. We recommend that you limit the ability to update or delete server metadata as follows:
• The SAS Administrators group should be able to administer, update, and delete all server definitions.
• Users who assign libraries, stored processes, or an OLAP schema to an application server must have WriteMetadata permission for that application server.
• No other users should be able to update or delete server definitions.
T I P
This chapter explains how to manage access to metadata objects that represent server processes. If you want to limit what a particular server process can do, see
“Locked-Down Servers” on page 96
instead.
Method
To avoid repeatedly adding the same explicit grants and denials to multiple pieces of server metadata, create and use an ACT. If you want to enable regular users to assign libraries, stored processes, or an OLAP schema to a particular application server, supplement the ACT settings with an explicit grant of WriteMetadata permission on that application server.
The following table depicts typical protections.
82
Chapter 7 • Access to Server Definitions
Table 7.1 Example: Protecting Server Definitions
Object
SASApp
Each logical server inside SASApp
SASMeta
Other immediate children of Server Manager
Direct Controls
ACTs Explicit Grants
Protect
Protect
DataAdmins: +WM
(none)
Protect
(none)
Protect
(none)
Note: The initial configuration in a new deployment limits access to the logical
workspace server and the logical SAS DATA step batch server within SASMeta, so it is not necessary to add protections to those components.
Instructions
Here is one way to set the permissions:
1. Log on to SAS Management Console as a member of the SAS Administrators group
(for example, sasadm@saspw). Select the Plug-ins tab.
2. (Optional) Examine the current settings.
a. Expand Server Manager, right-click the SASApp application server and select Properties.
b. On the Authorization tab, select SASUSERS. Notice that this group (which includes all registered users) has both ReadMetadata and WriteMetadata permissions. The application server inherits these grants from the standard repository-level settings. Click OK to close the dialog box.
3. If you do not already have an ACT with the appropriate pattern, create a new ACT as follows: a. Expand Authorization Manager, right-click Access Control Templates, and select New Access Control Template.
b. On the General tab, enter a name such as
Protect
.
c. On the Permission Pattern tab, define settings that you want to apply to all server metadata.
Table 7.2 Example: Pattern for the Protect ACT
Group
PUBLIC
Permission Pattern
Deny WriteMetadata, WriteMemberMetadata,
CheckInMetadata, Write, Administer
Group
SAS Administrators
Protect Server Definitions
83
Permission Pattern
Grant WriteMetadata, WriteMemberMetadata,
CheckInMetadata, Write, Administer,
ReadMetadata
*
*
These grants ensure that administrators can manage all metadata.
T I P
To increase reusability, this pattern includes permissions that are not relevant for server definitions.
d. On the Authorization tab, add explicit controls to protect the ACT that you are creating:
• Select PUBLIC and deny WriteMetadata permission. Leave the indirect
ReadMetadata setting in place.
• Select SAS Administrators and grant WriteMetadata permission. Leave the indirect ReadMetadata setting in place.
T I P
If the Users and Groups list box on the ACT's Authorization tab is empty, click OK to save the ACT. Then, right-click the new ACT, select
Properties, and select the Authorization tab again.
e. Click OK.
4. (Optional) If you want to allow some nonadministrators to assign libraries, stored processes, or an OLAP schema, and you do not already have a group that represents those users, create a new custom group. a. Right-click User Manager and select New
ð
Group.
b. On the General tab, enter a name such as
Data Admins
.
c. On the Members tab, move users (or groups) to the Selected Identities list box.
d. Click OK to save the new group. You will grant WriteMetadata permission to this group in step 5c. This group does not participate in the ACT's pattern because this group needs WriteMetadata permission on only some of the target objects.
Note: As an alternative to creating a group for only this purpose, you can skip this
step and instead assign the permissions to specific users in step 5c.
5. To set the protections: a. Expand Server Manager, right-click the first application server (if this is your first pass) and select Properties.
b. On the Authorization tab, click Access Control Templates. In the Add and
Remove Access Control Templates dialog box, move the
Protect
ACT to the
Currently Using list box (you have to expand the Foundation node to get to the
ACT). Click OK to return to the Authorization tab.
Note: Review the revised settings. Notice that SASUSERS is now denied
WriteMetadata permission and that PUBLIC and SAS Administrators have some ACT settings. These settings come from the
Protect
ACT.
c. (Optional) If the current item is an application server to which nonadministrators will assign libraries, OLAP schemas, or stored processes, click
Add, add one or more identities to the Authorization tab, and give each of those
84
Chapter 7 • Access to Server Definitions
identities an explicit grant of WriteMetadata permission. For example, you might assign the grant to a group (such as GroupA) or to individual users.
T I P
For any users that are under change management, grant CheckInMetadata
(CM), instead of WriteMetadata (WM). See SAS Intelligence Platform:
Desktop Application Administration Guide.
Note: Do not extend WriteMetadata access to the SASMeta application server,
because that server should be used for only a few specialized administrative tasks.
d. Click OK to save the settings for this object.
e. Repeat steps 5a-d for every immediate child of Server Manager. Immediate children include objects such as other application servers, third-party servers, the share server, the content server, and spawners.
f. Apply the
Protect
ACT to every logical server that is under an application server where you added an explicit grant of WriteMetadata permission (for
SASUSERS or a custom group such as
Data Admins
).
Note: This protects lower-level server metadata that would otherwise inherit the
nonadministrator's WriteMetadata grant from the application server. This also prevents nonadministrators who have WriteMetadata permission on the application server from deleting that application server.
See Also
“Use and Enforcement of Each Permission” on page 67
Hide Server Definitions
Introduction
In a new deployment, all registered users can see and use server definitions. You might choose to limit the availability of certain servers in any of the following circumstances:
• You want to create different levels of host access to data.
• You want to direct power users to a server with settings that offer advanced options.
• You want to direct high priority users to a server on hardware that offers superior performance.
• You want to enable users in SAS Information Map Studio to use a standard workspace server even when a logical pooled workspace server is present.
Method
If you choose to limit the availability of a server, preserve access as follows:
• Make sure that the SAS System Services group has ReadMetadata permission for server metadata. This enables the SAS Trusted User to see server definitions. This is necessary because the object spawner uses the SAS Trusted User to discover and read all server metadata.
Hide Server Definitions
85
Note: Users should not be members of the SAS System Services group. This group
is for service identities. In the standard configuration, the only member of this group is the SAS Trusted User.
• Make sure that the SAS General Servers group has ReadMetadata permission for server metadata. This enables the metadata identity of the launched server to see the server definition. This is a requirement for stored process servers and pooled workspace servers. This is not a requirement for standard workspace servers.
Note: Users should not be members of the SAS General Servers group. This group
is for service identities. In the standard configuration, the only member of this group is the SAS Trusted User.
• Metadata administrators should have ReadMetadata permission for all server metadata.
• Any user who will use a particular server needs ReadMetadata permission for that server, with the following exceptions:
• The requirement for ReadMetadata permission does not apply to requests to use a client-side pooled workspace server. A user can use a client-side pooled workspace server even if that user cannot see that server definition.
• The requirement for ReadMetadata permission is not enforced if the Use Server
Access Security check box on a logical server's Options tab is present and not selected. This check box should always be selected.
To efficiently set the permissions, create an ACT that includes the baseline grants and denials that you would use when you hide any server. To enable selected users to use a particular server, supplement the ACT settings with an explicit grant of ReadMetadata permission on that server.
For example, the following table summarizes settings that you might add to provide mutually exclusive access to two server components beneath a standard workspace server that is configured for SAS token authentication:
Table 7.3 Example: Hiding Server Definitions
Object
Direct Controls
*
ACTs Explicit Grants
SASApp - ServerA HideServer GroupA: +RM
SASApp - ServerB HideServer GroupB: +RM
*
The direct controls in this example do not determine which of the users who can see the server can also update or delete the server. See
“Protect Server Definitions” on page 81
.
Someone who has ReadMetadata permission for both ServerA and ServerB (for example, a member of the SAS Administrators group) uses the first server in the object spawner's list of servers.
Instructions
Here is one way to set the permissions:
1. Log on to SAS Management Console as a member of the SAS Administrators group
(for example, sasadm@saspw). Select the Plug-ins tab.
86
Chapter 7 • Access to Server Definitions
2. (Optional) To examine the current settings: a. Expand Server Manager, right-click the server or server component that you are limiting use of and select Properties.
Note: The SASMeta application server should have limited availability, because
it should be used only as instructed (in a few specialized administrative tasks).
b. On the Authorization tab, select SASUSERS. Notice that this group (which includes all registered users) has ReadMetadata permission. The application server inherits the grant from the standard repository-level settings. Click OK to close the dialog box.
3. To create the ACT: a. Expand Authorization Manager, right-click Access Control Templates, and select New Access Control Template.
b. On the General tab, enter a name such as
HideServer
.
c. On the Permission Pattern tab, define baseline settings for hiding servers.
Table 7.4 Example: Pattern for the HideServer ACT
Group
PUBLIC
SAS Administrators
SAS General Servers
SAS System Services
Permission Pattern
Deny ReadMetadata
Grant
ReadMetadata
*
Grant
Grant
ReadMetadata
ReadMetadata
**
*
**
This grant ensures that administrators can manage all server metadata (for alternatives, see
“Separated Administration” on page 78
).
This grant ensures that the SAS Trusted User (who is a member of SAS System Services) can read server metadata for the object spawner (on behalf of all users).
T I P
This pattern, when applied to a standard workspace server, grants a little more access than is strictly necessary. For a standard workspace server, the
SAS General Servers group does not need ReadMetadata permission. If you want to avoid this, consider omitting the SAS General Servers group from this ACT and remembering to add an explicit grant for this group when you are hiding a stored process server or pooled workspace server.
d. On the Authorization tab, protect the ACT that you are creating. Either apply an
ACT or add explicit controls that deny WriteMetadata permission to PUBLIC and grant WriteMetadata permission to the SAS Administrators group.
Note: If the Users and Groups list box on the ACT's Authorization tab is
empty, click OK to save the ACT. Then, right-click the new ACT, select
Properties, and select the Authorization tab again.
e. Click OK.
4. (Optional) If you do not already have a group that represents the users who will use the server, create a new custom group. a. Right-click User Manager and select New
ð
Group.
See Also
Hide Server Definitions
87
b. On the General tab, enter a name such as GroupA.
c. On the Members tab, move users (or groups) to the Selected Identities list box.
d. Click OK to save the new group. You will grant ReadMetadata permission to this group in step 5c. This group does not participate in the general pattern because this group does not need ReadMetadata permission on all servers.
Note: As an alternative to creating a group for only this purpose, you can skip this
step and instead assign the permissions directly to specific users in step 5c.
5. To set the permissions: a. Under Server Manager, right-click the server that you are limiting use of and select Properties.
b. On the Authorization tab, click Access Control Templates. In the Add and
Remove Access Control Templates dialog box, move the HideServer ACT to the
Currently Using list box (you have to expand the Foundation node to get to the
ACT). Click OK to return to the Authorization tab.
Note: If the Currently Using list already includes another ACT (such as the
Protect ACT), do not remove that assignment.
Note: Review the revised settings. Notice that SASUSERS is now denied
ReadMetadata permission and that PUBLIC and SAS Administrators have some ACT settings. These settings come from the
HideServer
ACT.
c. Click Add, add one or more identities to the Authorization tab, and give each of those identities an explicit grant of ReadMetadata permission. For example, you might assign the grant to a group (such as GroupA) or to individual users.
d. Click OK.
6. If you are limiting use of a logical server or server component, ensure that the Use
Server Access Security check box on the logical server's Options tab is selected. If the check box is not selected, requirements for ReadMetadata permission for that server and its components are not enforced. This option affects only enforcement of the ReadMetadata permission.
•
“Locked-Down Servers” on page 96
•
88
Chapter 7 • Access to Server Definitions
89
Chapter 8
Access to SAS Data
Stop Server Processes to Pick Up Lockdown Changes . . . . . . . . . . . . . . . . . . . . . 100
Using the Lockdown Feature with SAS Applications . . . . . . . . . . . . . . . . . . . . . . 100
Publishing Stored Processes to a Locked Down Server . . . . . . . . . . . . . . . . . . . . . 101
About Access to SAS Data
A SAS table (data set) is stored as a file in a host operating system. In general, a SAS table can be accessed and read by anyone who has the necessary host operating system permissions to that file. For this reason, it is essential to consider and address physicallayer access to SAS libraries and tables.
Here are some ways to protect SAS data:
• Provide only mediated access, so end users do not need or have physical access to the data. See
“Mediated Access” on page 90 .
• Bind the data to metadata, so all SAS requests for the data are subject to metadata
layer permissions. See “Metadata-Bound Libraries” on page 94 .
• Limit the reach of SAS servers in your deployment, so requests that are processed by those servers succeed only if the target host resources are included in a list of valid
(permitted) resources. See
“Locked-Down Servers” on page 96
.
90
Chapter 8 • Access to SAS Data
As an additional measure, you can encrypt SAS data at rest. See “ENCRYPT= Data Set
Option” in SAS Data Set Options: Reference .
Mediated Access
Overview
You can use the SAS metadata authorization layer to constrain access to SAS data.
However, metadata-layer constraints apply only when the user requests that data in a metadata-aware context.
As a simple example, assume that you have registered a library and several tables in the metadata. In SAS Management Console, you deny userA the ReadMetadata permission for a table named Salary. This prevents userA from seeing the Salary table in metadataaware applications. However, if userA has host access to the physical file that contains the table data, userA can open that file in Base SAS and examine all of the data. When the table is accessed directly, the metadata layer denial of ReadMetadata permission is not applied. In the following figure, the first image illustrates the issue and introduces the concept of mediated access.
Figure 8.1 Host Access to SAS Data
A user who has host access to SAS data can bypass metadata layer controls.
To prevent bypass, limit host access and provide only mediated access.
Metadata Layer Controls Metadata Layer Controls
Mediation
Privileged
Service Account
SAS Data
Host
Layer
Controls
SAS Data
With mediated access, a workspace server or stored process server retrieves SAS data from the host operating system using a service account called a launch credential. The server uses one account to retrieve data on behalf of each requesting user. Mediation is an integral part of a standard configuration as follows:
Mediated Access
91
• Host access from the stored process server and the pooled workspace server is always mediated.
• Host access from the standard workspace server is mediated if the server is configured to use SAS token authentication or if the server is providing client-side pooling.
Mediation offers the following advantages:
• Mediation makes pooling of workspace servers possible. For web applications, in particular SAS Web Report Studio, some form of pooling (client-side or server-side) is strongly recommended for performance reasons. Both forms of pooling always result in mediated host access to SAS data.
• Mediation facilitates partial access to a SAS table through metadata-layer controls
(for example, row-level or column-level access). Host-layer controls are at the file level. Such controls can grant or deny access to only an entire table.
• Mediation helps you avoid creating end-user accounts on back-end servers that host
SAS data.
CAUTION:
Along with its advantages, mediation introduces some risk.
If your deployment includes sensitive data in SAS tables, it is essential to review existing host-layer controls and to understand the effects of mediated access.
Considerations
Security considerations of mediated host access are as follows:
• Someone might obtain the service account ID and password and use those credentials to directly access the data.
• Someone might misuse the server in order to access data in a manner that circumvents metadata-layer controls.
The following table describes techniques for reducing risk. Keep in mind that there is no absolute security and that security goals must be balanced against other considerations.
Table 8.1 Alternatives for Reducing Mediation Risks
Modification Risk Reduction
Use a different launch credential for each server. For example, configure the serverside pooled workspace server with its own dedicated launch credential.
Diversification reduces exposure for each launch credential.
Limit the availability of SAS Information
Map Studio by installing that application only where it is needed.
Reduces opportunity to misuse that application.
Notes
Recommended if you have sensitive data in SAS tables.
Recommended if you have server-side pooling.
92
Chapter 8 • Access to SAS Data
Modification Risk Reduction
Limit the SAS folder locations from which relational information maps can be used.
Reduces opportunity to use a rogue information map, but affects only SAS Web
Report Studio and SAS Web Report Viewer.
Notes
This measure is necessary as one part of making row-level permissions that are defined in information maps fully secure. See
“Limit the Availability of
Relational Information
Maps That Implement
Row-Level Security” in
SAS Intelligence Platform:
Web Application
Administration Guide
.
Set up client-side pooling using a protected, dedicated service account as the puddle login.
Reduces opportunity to misuse the server by creating a rogue application. A client-pooled workspace server can be used from only designated web applications.
Delete the server-side pooled workspace server (and, for performance reasons, set up client-side pooling for SAS Web
Report Studio).
Eliminates opportunity to misuse that server
(in SAS Information Map Studio or in a rogue application).
Set up a restricted deployment of SAS
Web Report Studio with client-side pooling.
Fully isolates the restricted client-side pool
(with a dedicated pool administrator as well as a dedicated puddle login).
See “Secure Environment for BI Row-Level
Permissions” in SAS
Guide to BI Row-Level
Permissions.
Example
The preceding figure depicts a simple case where there is only one level of host access.
You can establish multiple levels of host access by setting up multiple servers. Each server has its own launch credential. This can involve a trade off of granularity ("how many different levels of access will I establish?") against administrative effort ("how much server metadata will I create and manage?").
For example, the following instructions show how you can physically isolate sensitive data that is accessed through a logical workspace server that uses SAS token authentication.
1. In the operating system that hosts the SAS data: a. Establish a separate file system directory for each level of access. This enables you to avoid setting host permissions on individual files. For example, to separate HR tables, you might create an HR directory.
b. Create or identify one service account for each level of access. For example, to provide physical protection for HR data, you might add an hrsrv account.
Note: For servers on Windows, each launch account must have the Windows
privilege Log on as a batch job.
c. Set operating system permissions so that each account has appropriate access to your SAS data. For example, deny the general use account (sassrv) physical
Mediated Access
93
access to the HR data and grant the HR account (hrsrv) physical access to all data
(including the HR data).
Note: If you want certain IT staff to be able to access the data directly using their
own accounts, preserve that access.
2. In SAS Management Console's User Manager, select the SAS General Servers group, right-click, and select the Accounts tab. For each service account that you created in step 1, click New and enter the service account credentials as an additional login for the SAS General Servers group. Include a password in each of these logins.
For example, you might enter the HR login as follows:
DefaultAuth | WIN\hrsrv | password
Note: Each of these logins will be directly associated with a particular server,
instead of being looked up by authentication domain. For this reason, it does not matter which authentication domain you use. You can ignore any warnings about the SAS General Servers group having multiple logins in the same authentication domain.
Note: Users should not be members of the SAS General Servers group. This group
is for service identities only. Users do not retrieve these credentials, so users should not have access to these logins. In a standard configuration, these logins are available only to the privileged service identity (the SAS Trusted User) that the spawner uses to read all server metadata.
3. In Server Manager, select the logical workspace server , right-click, and select
Properties. On the Options tab, verify that the Use server access security check box and the SAS token authentication radio button are selected. Click Cancel.
4. Select the logical workspace server again, right-click, and select Add Server. Set up one additional server for each level of access. All of the servers can use the same port. For example, you might add a server named Human Resources.
5. Beneath the logical workspace server, complete these steps for each server that you added: a. Select the server, right-click, select Properties, and select the Options tab. In the
Launch credentials drop-down list, select one of the service accounts (use a different account for each server). For example, for the Human Resources server, you would first select More logins and then search for and select the login that references the hrsrv account.
Note: In order to see this login in the list, you need user administration
capabilities.
b. On the Authorization tab, use the ReadMetadata permission to control who can use that server. Be sure to preserve necessary service access.
6. Under Server Manager, select the object spawner , right-click, and select
Properties. On the object spawner's Servers tab, add the new servers to the Selected
Servers list.
7. To make the changes take effect: a. Expand the object spawner , right-click the computer , and select Connect. b. Right-click the computer again and select Refresh Spawner. In the message box, click Yes.
c. Right-click the computer a third time and select Disconnect.
94
Chapter 8 • Access to SAS Data
8. To validate, expand the logical workspace server and select a server . On each connection object in the display area, right-click and select Test Connection.
Test the connections for every other sibling server that you added under the logical workspace server.
If you want to also enable power users to access data using their own accounts, set up a separate application server for those users. In that application server, add a logical workspace server that uses some form of host authentication. Even though regular users would not be able to authenticate (because they would not have host accounts for the workspace server machine), it is a good practice to limit ReadMetadata access to the additional application server.
You can use a similar approach to set up multiple levels of host access for a stored process server or a server-pooled workspace server. For multiple levels of host access from a client-pooled workspace server, set up multiple puddles. Always ensure that the server's host identity meets all of the requirements for a launch credential.
See Also
•
“Windows Privileges” on page 19
•
“Launch Credentials” on page 126
•
Appendix 4, “Choices in Workspace Server Pooling,” on page 263
Metadata-Bound Libraries
Overview
A metadata-bound library is a physical library that is tied to a corresponding metadata object.
Each physical table within a metadata-bound library has information in its header that points to a specific metadata object (a secured table object). The pointer creates a security binding between the physical table and the metadata object. The binding ensures that SAS universally enforces metadata-layer permission requirements for the physical table—regardless of how a user requests access from SAS.
Access from SAS to data within a metadata-bound library is provided only if all of the following conditions are met:
• The requesting user can connect to the metadata server in which the corresponding object is stored.
• The requesting user’s metadata identity has all required metadata-layer effective permissions for the requested action.
• The host identity with which the data is retrieved has all required host-layer access to the data.
Model
The following figure depicts the authorization model for a traditional table and a metadata-bound table. In both cases, UserA references the target data directly (for
Metadata-Bound Libraries
95
example, through a LIBNAME statement) and UserB requests the target data through a client that uses metadata to locate data (for example, SAS Web Report Studio).
Figure 8.2 Authorization Checks (by Data Type and Access Method)
Access to a Traditional Tabl e
UserA UserB
Access to a Metadata-Bound Table
UserA UserB
Metadata Layer Controls Metadata Layer Controls
= authorization check
(metadata layer)
= authorization check
(host layer)
Traditional Table Metadata-Bound Table
The preceding figure depicts the following key difference:
• When accessing a traditional table, a user can bypass metadata-layer controls by making a direct request.
• When accessing a metadata-bound table, a user cannot completely bypass metadatalayer controls. Even on a direct request, UserA is always subject to a metadata-layer permissions check before accessing SAS data from SAS.
For the metadata-bound table, the upwards-facing arrows are caused by the physical data’s security binding. For each metadata-bound table, information within the table header identifies a corresponding metadata object (a secured table object). Metadatalayer permissions on each secured table object affect access from SAS to the corresponding physical table.
For the metadata-bound table, UserB is subject to two metadata-layer authorization checks against two different metadata objects.
• The first check is against a traditional table object (for example, verifying that UserB has the ReadMetadata permission).
• The second check is against a secured table object (for example, verifying that UserB has the Select permission).
T I P
In the SAS metadata, traditional table objects and secured table objects are distinct and independent types of objects. See “Object Creation, Location, and
Inheritance” in SAS Guide to Metadata-Bound Libraries .
Here are some additional details about the preceding figure:
• The requesting users do not supply library or table passwords.
96
Chapter 8 • Access to SAS Data
• The metadata-layer authorization checks are against the metadata identity of the requesting user. The host-layer authorization checks are against the identity of the
SAS process that retrieves the data.
• The figure addresses access to SAS data from SAS, not interaction through host commands.
• The figure is conceptual, simplified, and abstracted. It is not intended as a detailed technical specification.
Encryption Features
Metadata-bound tables can be encrypted in the same way as traditional tables, except that the Read password is obtained from metadata instead of being supplied in SAS code. Beginning in the first maintenance release for SAS 9.4, you can require encryption for the tables in a metadata-bound library. In addition, a metadata-bound library’s AES encryption key can be stored in metadata so that the user does not have to provide the encryption key to access the data set.
See Also
SAS Guide to Metadata-Bound Libraries
Locked-Down Servers
Overview
You can limit the reach and activities of a SAS server by putting it in a locked-down state. This feature supplements the standard operating system-level and metadata-layer protections by giving SAS processes access to only specific physical resources on the server. This feature is intended for sites that need an extra level of security. For example, you might want to physically separate sensitive data from other data by placing the sensitive data on a separate, locked-down server.
The following types of servers can be locked down:
• workspace (standard)
• pooled workspace (server-side pooling)
• stored process
• batch
• grid
• SAS/CONNECT
The lockdown feature is not supported for SAS/SHARE servers. The feature is also not supported for certain SAS products and solutions. For more information, see SAS Note
51644 .
For details about how a particular solution uses lockdown, consult the administration guide for that solution.
Locked-Down Servers
97
The Lockdown Point
During initialization of the SAS environment, all paths and files for which the operating system allows access are available. Initialization of the SAS environment includes metadata library pre-assignments and autoexec execution (which might include library pre-assignments). Having all paths and files available during initialization facilitates one-time setup of library and file assignments so that they function as usual.
When initialization is complete, SAS enters the lockdown state, which is described in the next topic.
The Lockdown State
Resources That Are Accessible to Locked-Down Servers
When a server has entered the lockdown state, the server can access only the following host resources:
• pre-assigned libraries that are defined in metadata, as well as libraries and external files that are assigned in autoexec files or through INITSTMT execution. Users can access these resources only by using their assigned librefs and filerefs.
• host paths and files that are included in the server’s list of permitted resources. This list is referred to as the lockdown path list. The lockdown path list is established during server initialization and is finalized at the lockdown point. The list includes the following:
• paths and files that you specify using a LOCKDOWN statement. You can include one or more of these statements in the autoexec file for the server. Users can access these resources (including subdirectories within the specified paths) directly by name, without relying on a libref or fileref.
• resources that are essential for server processing, such as system paths that are defined in SAS configuration files and paths of stored process repositories that are defined in metadata. These paths are implicitly included in the lockdown path list and do not need to be specified in the LOCKDOWN statement.
• for servers executing in the z/OS environment, temporary z/OS data sets that are created on behalf of clients.
T I P
The lockdown path list is a whitelist. That is, it specifies which paths are accessible. It does not specify which paths are inaccessible.
For example, a user who connects to a locked-down workspace server from SAS
Enterprise Guide can access only those host paths and files that are included in that server’s list of permitted resources. Regardless of host-layer permissions, LIBNAME and FILENAME statements that users submit to a locked-down server are rejected, unless the target resource is included in the server’s lockdown path list.
Note: For details about the LOCKDOWN statement and the contents of the lockdown
path list, see “LOCKDOWN Statement” in SAS Intelligence Platform: Application
Server Administration Guide
.
Note: When you make a change that affects the lockdown path list (for example, when
you modify the LOCKDOWN statement or define a new stored process repository),
be sure to restart your servers to pick up the changes .
98
Chapter 8 • Access to SAS Data
SAS Language Constraints for Locked-Down Servers
A server that is in the lockdown state is constrained as follows:
• The server cannot run DATA step Java object methods.
• The server cannot run the JAVAINFO procedure and the GROOVY procedure.
• The server cannot invoke the following DATA step functions: ADDR, ADDRLONG,
MODULE, PEEK, PEEKC, PEEKCLONG, PEEKLONG, POKE, and POKELONG
• A server running in the z/OS environment cannot perform the following functions:
• run the PDS, PDSCOPY, RELEASE, SOURCE, TAPECOPY, and TAPELABEL procedures
• execute any of the DATA step functions that are part of the SAS Data Location
Assist feature. For details, see SAS Companion for z/OS.
• use the Volume Table of Contents (VTOC) access method
• For a server running in the z/OS environment, the DSNEXST statement returns a value of false for a data set that is not in the lockdown path list.
Restricted Access Methods and Procedures
By default, when SAS is in a locked-down state, the following access methods are not available: EMAIL, SOCKET, TCPIP, URL, HTTP, FTP, and HADOOP. As a result, the
HTTP, SOAP, and HADOOP procedures are also not available. These methods and procedures are initially disabled, keeping users from potentially generating spam or accessing files in a server that is not configured for authentication. As an administrator, you can enable these methods and procedures, allowing them to be accessible to users.
One or more of the access method names can be specified in the following
LOCKDOWN statement:
LOCKDOWN ENABLE_AMS = access-method-1 <access-method-2...>;
Alias name pairs exist for certain access methods. If one of the access methods in the pair is enabled, the other access method is automatically enabled. Here is a list of the available alias name pairs:
• URL and HTTP
• SOCKET and TCPIP
In addition, if either URL or HTTP is enabled, the HTTP and SOAP procedures are automatically enabled. If HADOOP is enabled, HADOOP procedure is also enabled.
For additional details and examples about this statement, see “LOCKDOWN Statement” in SAS Intelligence Platform: Application Server Administration Guide .
Best Practices for Using the Lockdown Feature
When using the lockdown feature, it is strongly recommended that you follow these best practices:
• Make sure that physical resources on the locked-down server are protected by appropriate operating system permissions. The lockdown feature is not intended to replace operating system protections.
• Make sure that the NOXCMD system option has been enabled on the locked-down server.
Locked-Down Servers
99
• Deny most users WriteMetadata access to the server context (for example, SASApp).
Users might need WriteMetadata access to the server context for the following reasons:
• to modify server definitions in metadata (including start-up options, ports, paths, and access controls).
• to add or remove library definitions on the server.
• to create source code repositories on the server’s file system and to publish stored processes to these repositories. For more information, see
Processes to a Locked Down Server” on page 101
.
The Audit.Lockdown Logger
If a user attempts to access a path that is not available, the following error is issued to the
SAS log and to the Audit.Lockdown logger (if the SAS session is configured to use the
SAS logging facility).
ERROR: The path <path> is invalid because it is not in the list of accessible paths when SAS is in the lockdown state.
The following errors are issued to the SAS log, in each of the specified situations.
• If a user attempts to access a method that is not available:
ERROR: The access method "<name>" is disabled when SAS is in the lockdown state.
• If a user attempts to access a procedure that is not available:
ERROR: The PROCEDURE <name> is disabled when SAS is in the lockdown state.
How to Lock Down a SAS Server
Lock Down a Specific SAS Server
To lock down a specific SAS server (for example, a workspace server), follow these steps:
1. Add the LOCKDOWN system option to the sasv9_usermods.cfg file that is located in the server’s configuration directory.
2. Add a LOCKDOWN global statement to the autoexec_usermods.sas file that is located in the server’s configuration directory. Use the statement’s PATH= argument to specify which directories and files are available to the server’s users.
Alternatively, you can use the FILE= argument to specify the name of a file that contains a list of valid directories and files.
Here are examples of typical LOCKDOWN statements for a workspace server.
These statements provide users with access only to their home directories:
• Windows:
lockdown path="?FOLDERID_Profile";
• UNIX:
lockdown path ="~";
• zOS: You can enable access to a user’s personal z/OS data sets (typically cataloged under a high-level qualifier matching the user’s account), to the user’s
UNIX home directory, or to both, as shown in the following example:
lockdown path="~/"; lockdown path="mvs:.*";
100
Chapter 8 • Access to SAS Data
Note: When using the FILE= argument, if there are DBCS characters in the path list in
the lockdown text file, add the ENCODING option to the FILENAME statement.
This option ensures that the path list is transcoded properly to SAS session encoding.
Note: When using the FILE= argument, you can limit users’ access to the lockdown
path list by placing the file in a path that is not in the lockdown path list.
For additional details and examples, see “LOCKDOWN Statement” in SAS Intelligence
Platform: Application Server Administration Guide
.
Lock Down a SAS Application Server Context
For simplicity, you can lock down an entire SAS Application Server context so that the constraints apply to all servers in that context. Follow these steps:
1. Place the LOCKDOWN system option in the sasv9_usermods.cfg file that is located in the application server’s configuration directory (for example,
/Lev1/SASApp/ sasv9_usermods.cfg
).
2. Add a LOCKDOWN statement to your equivalent of the following file:
/Lev1/
SASApp/appserver_autoexec_usermods.sas
. In the FILE= or PATH= argument, specify which host resources are available to the servers in the context.
Here is an example of a LOCKDOWN statement for typical application server usage: lockdown path="~" "/SASData/reportdata" "/SASData/STPData" "/SASData/sales";
As an alternative, you can place LOCKDOWN statements in the autoexec_usermods.sas files for each of the servers in the context. This alternative is useful if you want a resource to be available only to a specific server in the context.
For example, you could place the following statements in the autoexec_usermods.sas files for the specified servers:
• workspace server:
lockdown path="user-home-directory";
(using the appropriate syntax for your operating system)
• stored process server:
lockdown path="/SASData/STPData/";
• pooled workspace server:
lockdown path="/SASData/ReportData/";
Stop Server Processes to Pick Up Lockdown Changes
When you make a change that affects the lockdown path list (for example, when you modify the LOCKDOWN statement or define a new stored process repository), the change does not affect currently running servers. After making changes, be sure to stop or quiesce any currently running processes on the affected pooled workspace server or stored process server. See “Use SAS Management Console to Control Processes on
Servers” in SAS Intelligence Platform: System Administration Guide . The changes will take effect the next time a server session is started.
Using the Lockdown Feature with SAS Applications
Before SAS client applications attempt to access the file system on a locked-down server, make sure the application’s default directory for server navigation is set to a location that is included in the lockdown path list. For example, do the following to modify the default starting path for SAS Enterprise Guide:
1. In SAS Management Console, open Server Manager and navigate to the appropriate application server (for example, SASApp) and logical workspace server.
Locked-Down Servers
101
2. Under the logical workspace server node, right-click the node for the workspace server.
3. Select Properties
ð
Options
ð
Advanced Options
ð
File Navigation.
4. In Path, specify a path that is included in the server’s lockdown path list.
Some SAS applications might require different procedures for accessing locked-down servers. For more information, consult SAS Notes or the administration documentation for the application.
T I P
The lockdown feature is not supported for certain SAS products and solutions.
For more information, see SAS Note 51644 .
Publishing Stored Processes to a Locked Down Server
If users need to be able to publish stored processes to a locked-down server, it is important to consider the effect of this activity on the lockdown path list. Users can publish stored processes using two different methods:
• by keeping stored process code in source code repositories in the server’s file system.
To use this method, the user must have WriteMetadata permission on the server context (for example, SASApp). Users who have WriteMetadata access to the server context can also define new source code repositories. When a server process is initiated, all stored process repository paths that have been defined in metadata for that server are implicitly added to the server’s lockdown path list. Therefore,
WriteMetadata access should be provided to only a limited number of users.
• by keeping stored process source code in metadata instead of in the server’s file system. To use this method, the user does not need WriteMetadata access to the server context. Since source code repositories do not need to be defined, the server’s lockdown path list is not affected.
See Also
• “LOCKDOWN System Option” in SAS Intelligence Platform: Application Server
Administration Guide
• “LOCKDOWN Statement” in SAS Intelligence Platform: Application Server
Administration Guide
102
Chapter 8 • Access to SAS Data
103
Chapter 9
Batch Tools for Metadata
Authorization
About the Batch Tools for Metadata Authorization . . . . . . . . . . . . . . . . . . . . . . . .
About the Batch Tools for Metadata Authorization
Introduction
The batch tools for access control are Java scripts that enable you to connect to a metadata server and perform the following tasks:
• view, set or remove direct access controls (explicit grants, explicit denials, permission conditions, and directly applied access control templates) for a specified object
• create a new access control template, update an existing access control template, and display an ACT’s permission pattern
Batch utilities for metadata-layer access control are located in your equivalent of the following path:
SASHome\SASPlatformObjectFramework\9.4\Tools
.
104
Chapter 9 • Batch Tools for Metadata Authorization
General Options
Table 9.1 Connection Options
Option
-host host-name
-password password
-port port
Description
Identifies the host of the metadata server. This option is required if the
-profile
option is not set.
Specifies the password of the connecting user. This option is required if the
-profile
option is not set.
Specifies the port on which the metadata server runs. This option is required if the
-profile
option is not set.
-profile profile-name Specifies a connection profile. You can use this option instead of using individual options (
-host
,
-port
,
-user
, and
-password
) to provide connection information.
*
-user user-ID Specifies the user ID of the connecting user. This option is required if the
-profile
option is not set.
*
The connection profile must exist on the same physical machine where you are running the tools. See
“Administering Connection Profiles for Desktop Applications” in SAS Intelligence Platform: Desktop
Application Administration Guide
.
Table 9.2 Logging and Help Options
Option
-log log-file
-?
Description
Identifies a log file or directory.
Displays online Help.
Tips and Details
Specifying Objects
You can specify any public object that can be identified by its metadata location. If you specify an object other than a folder, you must specify the object type in parentheses after the object name. Object types are case sensitive. For a list, see “Specifying Object
Types” in SAS Intelligence Platform: System Administration Guide .
Note: In the current release, specifying multiple objects (by using wildcards or filter
criteria) is not supported.
Specifying Permissions
You can use either full permission names (such as ReadMetadata) or permission abbreviations (such as RM) to specify permissions. See
.
Batch Tool: sas-show-metadata-access
105
Specifying Identities
When you specify an identity name, the tool looks first for a user with that name. If a user is not found, the tool looks for a user group with that name. To avoid any possible ambiguity, you can append the object type to the name (for example, "sasdemo(User)" or
"Developers(UserGroup)."
Quotation Marks and Line Breaks
Enclose values that contain spaces or special characters in quotation marks.
Any line breaks in example code are for clarity in reading the documentation. Line breaks are not used in the batch utility commands.
More Information
For general information about administrative batch tools and details about other system administration tasks that you can perform in batch, see
SAS Intelligence Platform:
System Administration Guide
.
Batch Tool: sas-show-metadata-access
Syntax
The general form of the command is as follows:
sas-show-metadata-access connection-options object-path <special-options>
The following list provides details:
connection-options
provide connection information for the metadata server (see Table 9.1 on page 104 ).
object-path
specifies the metadata location (path) and name of the target object. If you specify an object other than a folder, you must specify the object type in parentheses after the object name.
<special-options>
specify the type and scope of access information to be displayed. See the following table.
Table 9.3 Special Options for sas-show-metadata-access
Option
-effective
-onlyGroup group-name
-onlyUser user-name
Description
Displays effective access. If this option is not set, only direct access controls are displayed.
*
Displays access for only the specified user group.
Displays access for only the specified user.
*
Effective access is the net effect of all applicable permission settings (both direct access controls and inherited permissions).
106
Chapter 9 • Batch Tools for Metadata Authorization
Examples
For the
DemoBranch/DemoFolder
object, show all direct access controls.
sas-show-metadata-access -profile Admin /DemoBranch/DemoFolder
Here is some example output from the preceding command:
-grant sasdemo:Read,ReadMetadata
-grant "SASAdministrators(UserGroup)":ReadMetadata
-deny "PUBLIC(UserGroup)":ReadMetadata
-grant "SAS System Services(UserGroup)":ReadMetadata
For the
DemoBranch/DemoFolder/DemoTable
object, show all direct access controls that are assigned to the SAS Demo User.
sas-show-metadata-access -profile Admin /DemoBranch/DemoFolder/DemoTable(Table)
-onlyUser sasdemo
Here is some example output from the preceding command:
-grant sasdemo:Read,ReadMetadata
For the
Shared Data
folder object, show SAS Demo User’s effective access.
sas-show-metadata-access -profile Admin "Shared Data"
-effective -onlyUser sasdemo
Here is some example output from the preceding command:
-grant sasdemo:ReadMetadata
-deny sasdemo:Administer,CheckInMetadata,Delete,
Read,Write,WriteMetadata,WriteMemberMetadata,Create
Batch Tool: sas-set-metadata-access
Syntax
The general form of the command is as follows:
sas-set-metadata-access connection-options object-path special-options
The following list provides details:
connection-options
provide connection information for the metadata server (see Table 9.1 on page 104 ).
object-path
specifies the metadata location (path) and name of the target object. If you specify an object other than a folder, you must specify the object type in parentheses after the object name.
special-options
provide details about the permissions that you want to set. See the following table.
Batch Tool: sas-set-metadata-access
107
Table 9.4 Special Options for sas-set-metadata-access
Option
-addACT act-name
-condition condition-expression
-deny identity-name:permission1<,permission2...>
-grant identity-name:permission1<,permission2...>
-remove identity-name:ALL
-remove identity-
name:permission1<,permission2...>
-removeACT act-name
-removeAll
Description
Directly applies the specified access control template.
**
Adds the specified expression
(as a permission condition) to the immediately preceding explicit grant.
*
For the specified user or group, sets explicit denials of one or more permissions.
For the specified user or group, sets explicit grants of one or more permissions.
For the specified user or group, removes all explicit grants or denials.
For the specified user or group, removes explicit grants or denials of one or more permissions.
Removes (de-applies) the specified access control template.
**
Removes all direct access controls (explicit grants, explicit denials, permission conditions, and directly applied
ACTs).
*
**
For information about permission conditions, see “Fine-Grained Controls for Data” on page 62 .
Permission conditions for information maps are not supported by the batch tool.
To reference an ACT that is stored in a custom repository, use the format "repository-name/template-
name".
Examples
On the
DemoBranch/DemoFolder/DemoTable
object, add explicit grants of the
Read and ReadMetadata permissions for the SAS Demo User.
sas-set-metadata-access -profile Admin /DemoBranch/DemoFolder/DemoTable(Table)
-grant sasdemo:Read,ReadMetadata
On the
DemoBranch/DemoFolder/DemoTable
object, replace explicit grants of the
Read and ReadMetadata permissions for the SAS Demo User with explicit denials.
sas-set-metadata-access -profile Admin /DemoBranch/DemoFolder/DemoTable(Table)
108
Chapter 9 • Batch Tools for Metadata Authorization
-deny sasdemo:Read,ReadMetadata
On the
DemoBranch/DemoFolder
object, apply the testACT.
sas-set-metadata-access -profile Admin /DemoBranch/DemoFolder
-addACT testACT
On the
DemoBranch/DemoFolder
object, remove the testACT.
sas-set-metadata-access -profile Admin /DemoBranch/DemoFolder
-removeACT testACT
On the
DemoBranch/DemoFolder
object, remove all direct access controls.
sas-set-metadata-access -profile Admin /DemoBranch/DemoFolder
-removeAll
On the
DemoBranch/DemoFolder
object, remove any existing direct access controls and add new explicit settings of the ReadMetadata permission to block access for all restricted users other than members of the SASAdministrators group.
sas-set-metadata-access -profile Admin /DemoBranch/DemoFolder
-removeAll -deny PUBLIC:RM -grant SASAdministrators:RM
On the
DemoBranch/DemoCube/Region
object, add a permission condition that enables the SAS Demo User to see only the North members within the Region dimension.
sas-set-metadata-access -profile Admin /DemoBranch/DemoCube/Region(Dimension)
-grant sasdemo:Read -condition "Region=North"
Batch Tool: sas-make-act
Syntax
The general form of the command is as follows:
sas-make-act connection-options act-name -create | -update | -show special-options
The following list provides details:
connection-options
provide connection information for the metadata server (see Table 9.1 on page 104 ).
act-name
specifies the name of the target (or new) ACT.
Note: To reference a template that is stored in a custom repository, use the format
"repository-name/template-name".
-create
creates a new access control template.
-update
updates an existing access control template.
-show
displays the permission pattern of an access control template.
Batch Tool: sas-make-act
109
special-options
specify the permission patterns that are to be added to a new or existing ACT, or specify that all permission patterns are to be removed from an existing ACT. See the following table.
Table 9.5 Special Options for sas-make-act
Option Description
-deny identity-name:permission1<,permission2...> For the specified user or group, adds denials of one or more permissions to the permission pattern of the new ACT.
-grant identity-
name:permission1<,permission2...>
-remove identity-name:ALL
For the specified user or group, adds grants of one or more permissions to the permission pattern of the new ACT.
For the specified user or group, removes all explicit grants or denials.
-remove identity-
name:permission1<,permission2...>
For the specified user or group, removes explicit grants or denials of one or more permissions.
-removeAll
Removes all settings from the permission pattern.
Examples
Display the permission pattern of the repository ACT (for the foundation repository).
sas-make-act -profile Admin "Default ACT" -show
Here is some example output from the preceding command:
-grant "SASAdministrators(UserGroup)":ReadMetadata,WriteMetadata,
CheckInMetadata,Administer
-grant "SASUSERS(UserGroup)":ReadMetadata,WriteMetadata,CheckInMetadata
-deny "PUBLIC(UserGroup)":ReadMetadata,WriteMetadata,WriteMemberMetadata,
CheckInMetadata,Read,Write,Administer,Create,Delete,Execute,
Create Table,Drop Table,Alter Table,Select,Insert,Update
-grant "SAS System Services(UserGroup)":ReadMetadata,WriteMetadata
In the permission pattern of the repository ACT (for the foundation repository), add grants of the Read and Write permissions for a group named Dev.
sas-make-act -profile Admin "Default ACT" -update -grant Dev:Read,Write
Create a custom ACT called Hide that blocks the ReadMetadata permission for everyone except the SAS Administrators and SAS System Services groups.
sas-make-act -profile Admin Hide -create -deny PUBLIC:RM
-grant "SASAdministrators":RM -grant "SAS System Services":RM
110
Chapter 9 • Batch Tools for Metadata Authorization
T I P
ACTs are not automatically protected against changes or deletion. Set permissions on each new ACT (either using the Authorization tab in SAS
Management Console or using the sas-set-metadata-access batch tool).
111
Part 3
Authentication
Chapter 10
Authentication Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Chapter 11
Authentication Mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Chapter 12
Authentication Tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
112
113
Chapter 10
Authentication Model
Authentication to Data Servers and Processing Servers . . . . . . . . . . . . . . . . . . . . .
Solution to Mixed Providers: Use SAS Token Authentication . . . . . . . . . . . . . . . . 118
Solution to Mixed Providers: Store User IDs and Passwords . . . . . . . . . . . . . . . . 119
How Many Authentication Domains Do I Need? . . . . . . . . . . . . . . . . . . . . . . . . . . 123
Outbound and Trusted Authentication Domains . . . . . . . . . . . . . . . . . . . . . . . . . . 124
How to Create and Designate a New Launch Credential . . . . . . . . . . . . . . . . . . . . 128
Who Can Launch a Standard Workspace Server? . . . . . . . . . . . . . . . . . . . . . . . . . 129
PUBLIC Access and Anonymous Access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
About PUBLIC Access and Anonymous Access . . . . . . . . . . . . . . . . . . . . . . . . . . 130
Introduction to the Authentication Model
There is no single mechanism that is applicable to all authentication events throughout a typical deployment. Each deployment uses some combination of authentication
114
Chapter 10 • Authentication Model
processes, trust relationships, and single sign-on technologies. This helps balance a range of goals such as the following:
• preserve individual identity
• minimize security exposures
• provide a unified user experience
• minimize set up and maintenance efforts
• provide access to disparate systems within an environment
• integrate into a wide variety of general computing environments
See Also
“How to Facilitate Authentication” on page 156
Authentication to the Metadata Server
How SAS Identity Is Determined
When a user launches a SAS client, the following process occurs:
1. In the verification phase, the system ensures that the user is who he or she claims to be. For example, this credential-based host authentication method might be used: a. The client prompts the user for an ID and password.
b. The user enters credentials that are known to the metadata server's host. c. The client sends the credentials to the metadata server.
d. The metadata server passes the credentials to its host for authentication.
e. If the host determines that the user has a valid account, the host returns the authenticated user ID to the metadata server.
2. In the SAS identity phase, the system resolves the authenticated user ID to a particular SAS identity. In this phase, SAS examines its copies of user IDs in an attempt to find one that matches the authenticated user ID. One of the following outcomes occurs:
• A matching user ID is found, so a connection is established under the owning identity. The owning identity is the user or group whose definition includes a login with the matching user ID.
Note: Integrity constraints ensure that there will not be more than one owning
identity. See
“Uniqueness Requirements” on page 22 .
Note: Not all applications allow a group identity to log on.
• No matching user ID is found, so a connection is established under the generic
PUBLIC identity. In the metadata layer, the user is a PUBLIC-only user.
Note: The matching process expects the SAS copy of the user ID to be qualified
(if it is a Windows user ID).
Note: Not all applications allow a PUBLIC-only user to log on. See
Access and Anonymous Access” on page 130 .
Authentication to the Metadata Server
115
Authentication Process and Methods
The following figures introduce the metadata server's authentication process and methods. In the following figures, notice these points:
• Only the verification phase varies; the SAS identity phase is always the same. With any approach, you need a well-formed user definition for each user who is not a
PUBLIC-only identity.
• Except where internal accounts are used, the process always involves two sets of identity information, one in an external provider and another in the metadata.
The following figure depicts the basic process.
Figure 10.1 Metadata Server: Host Authentication (Credential-Based)
The following figure depicts a special case where a metadata administrator named Joe uses an internal account.
116
Chapter 10 • Authentication Model
Figure 10.2 Metadata Server: Internal Authentication
1 joe@saspw
& password
2 joe@saspw
& password
4 accept
Client
Metadata
Server
login login login login login login
account
Joe
login
account sastrust account sasadm
3
Joe
Metadata Repository
For alternate approaches, see “Integrated Windows Authentication” on page 139
,
, and “Summary of Methods for LDAP Integration” on page 151
.
Authentication to Data Servers and Processing
Servers
A registered user who has a connection to the metadata server accesses the OLAP server, stored process server, and pooled workspace server seamlessly by SAS token authentication. Authentication to the following servers requires coordination: standard workspace server
To provide seamless access, coordinate the workspace server with the metadata server.
third-party data server (for example, Oracle)
To provide access, select an approach from the following table.
Table 10.1 Authentication to a Third-Party Data Server
Goal Approach
Provide seamless access and preserve individual identity to the target server.
Store individual user IDs and passwords in the metadata (each on the Accounts tab of a different user definition).
Provide seamless access.
Provide seamless access with a few distinct access levels for resources in the target server.
Store the user ID and password for one shared account in the metadata (on the Accounts tab of a group definition).
Store a few shared user IDs and passwords in the metadata (each in the Accounts area of a different group definition). Or, for a hybrid approach, use a combination of personal and group logins.
See Also
Goal
Preserve individual identity.
Mixed Providers
117
Approach
No configuration is required. Users are prompted for credentials for the target server. Secondary prompting is supported for desktop applications and SAS Web
Report Studio.
Note: In general, requesting users can access only those servers for which they have the
ReadMetadata permission. An exception is client-side pooling, in which access depends on membership in a puddle group.
Note: This topic is about metadata-aware connections. Direct connections to a SAS
server cannot use SAS token authentication. Direct connections use client-supplied credentials or, in some cases, Integrated Windows authentication.
•
“Coordinate the Workspace Server” on page 157
•
“How to Store Passwords for a Third-Party Server” on page 175
•
“SAS Token Authentication” on page 145
•
Appendix 4, “Choices in Workspace Server Pooling,” on page 263
Mixed Providers
About Mixed Providers
If the credentials with which users initially log on are not also valid for the workspace server, access to the standard workspace server is not seamless. For example, if you use an Active Directory account in an initial logon to a metadata server on Windows, and you attempt to access a standard workspace server on UNIX, you might be prompted for a user ID and password that are valid for the UNIX host.
Note: In general, web applications and SAS Information Map Studio use the pooled
workspace server, not the standard workspace server. So users who use only these applications might not be affected by a mixed provider situation.
In a mixed providers situation, achieving seamless access to the standard workspace server requires a trade-off between configuration effort, maintenance effort, and degree of segregation for host access from that server to SAS data. The following table outlines the choices.
Table 10.2 Choices for Seamless Access in a Mixed Provider Scenario
Priority Recommendation
Preserve each user's identity.
Align authentication so that both servers use the same provider.
Minimize administrative effort.
Convert the workspace server to use SAS token authentication.
118
Chapter 10 • Authentication Model
The following topics provide details about each approach.
Solution to Mixed Providers: Align Authentication
If you can enable the metadata server and the standard workspace server to use the same provider, access to the workspace server is seamless. Here are some examples:
• If the metadata server is on Windows and the workspace server is on UNIX, you might extend the UNIX host authentication process to recognize Windows accounts
(using a technology such as PAM). With this approach, the Windows credentials that users supply in their initial logon to the metadata server can be reused for authentication to the workspace server. User identity is preserved from the workspace server to the host, you do not have to store user passwords in the metadata, and access is seamless.
Note: If you chose to use IWA in this scenario, you should configure both servers to
support that authentication method. If you configure only the metadata server to support IWA, then access to the workspace server is not seamless, because there are no cached credentials available for reuse.
Note: This approach usually requires that you store two logins for anyone who
accesses the standard workspace server. One login contains the user's ID in qualified format (for example, WIN\myID) and the other login contains the same
ID in short format (for example, myID). Both logins can be in DefaultAuth.
Neither login has to include a password. An exception to this requirement is if the user names on UNIX are domain qualified (in that case, only the domain qualified user ID is stored in metadata).
• If the metadata server is on UNIX (or z/OS) and the workspace server is on
Windows, an additional alternative is to configure direct LDAP, so the metadata server itself recognizes Windows accounts.
Note: Direct LDAP is not supported for the workspace server. The workspace server
can use only some form of host authentication or SAS token authentication.
Solution to Mixed Providers: Use SAS Token Authentication
If you configure the workspace server to use SAS token authentication, access is seamless because the workspace server's host operating system is no longer used to validate users. Instead, users are validated by the metadata server through a single-use
SAS identity token. If you need to provide a few distinct levels of host-layer access, you can define a few servers, with each server using SAS token authentication and running under a distinct launch credential.
In a mixed providers situation, using SAS token authentication is preferable to using a group login for these reasons:
• SAS token authentication enables the requesting user's SAS identity to be used for metadata layer evaluations such as library pre-assignment and permissions and auditing.
• SAS token authentication does not give users direct access to the server launch credential.
• SAS token authentication involves less transmission of reusable credentials (the client does not retrieve credentials from the metadata and send those credentials to the workspace server).
Credential Gaps
119
Solution to Mixed Providers: Store User IDs and Passwords
Note: As explained in the preceding topic, storing credentials is not a first choice
solution to a mixed providers situation.
If you store user or group passwords for the workspace server's host, access is seamless through credential retrieval. In this solution, you treat the workspace server as if it were a third-party server such as Oracle.
See Also
•
“How to Configure SAS Token Authentication” on page 157
•
“Pluggable Authentication Modules (PAM)” on page 142
•
“Direct LDAP Authentication” on page 136
•
“How to Store Passwords for the Workspace Server” on page 174
Credential Gaps
A credential gap is a situation in which a user does not seamlessly access the workspace server for any of these reasons:
• Server configurations are incorrect, incomplete, or incompatible.
• The user's context does not include credentials that are known to the workspace server's host.
• The user's context does not pair credentials that are known to the workspace server's host with the workspace server's authentication domain.
• The workspace server is on Windows, using credential-based authentication and the user's host account does not have the Windows privilege Log on as a batch job.
The usual symptom of a credential gap is a prompt for a user ID and password after a user makes a request that requires a workspace server. A credential gap can be problematic for these reasons:
• The prompts interrupt the user experience.
• Users have to know credentials that are valid for the workspace server's host and know that those are the correct credentials to provide.
• Not all middle-tier services and web applications prompt for credentials (and, without a prompt, the user request fails).
You can use the following list to help troubleshoot a credential gap.
Troubleshooting Credential Gaps
If the user initially logs on via web authentication
The user's initial logon does not add a password to the user's context. Make sure that the web application uses some form of pooling. If the problem persists, consider configuring the workspace server to use SAS token authentication.
If the user initially logs on via Integrated Windows authentication
120
Chapter 10 • Authentication Model
The user's initial logon does not add a password to the user's context. Configure the workspace server for Integrated Windows authentication (IWA).
If the user logs on with a user ID that ends in @saspw
Tell the user that they get the prompt because they are using an internal account. When the user gets the additional prompt, they must enter a user ID and password that are known to the workspace server's host. The host account must correspond to a metadata identity that has ReadMetadata permission for the server definition. On Windows, the host account must have the Log on as a batch job privilege.
If the user's connection profile contains an @saspw user ID
The user's context does not pair the credentials from the user's initial logon with the
DefaultAuth authentication domain. Tell the user to create a new connection profile with external credentials (and no value in the Authentication domain field) and try again. To ensure optimal credential reuse, users should not use the same connection profile for both internal and external accounts.
If the user's connection profile has a value other than DefaultAuth for the authentication domain
The user's context does not pair the credentials from the user's initial logon with the
DefaultAuth authentication domain. Tell the user to either clear this field or enter the value
DefaultAuth
and try again.
If the user is in SAS Enterprise Guide and accessing a workspace server that is set to prompt
Verify the logical workspace server’s Options settings. If the setting is intentional, tell the user to supply host credentials.
If the workspace server is not assigned to the correct authentication domain
Credential reuse might be impaired. In most configurations, the workspace server should be in DefaultAuth. To verify (and, if necessary, correct) the workspace server's authentication domain assignment, select the Plug-ins tab in SAS Management Console, navigate to the server , select its connection object , right-click, and select
Properties. The authentication domain assignment is on the Options tab.
See Also
•
“Windows Privileges” on page 19
•
“Credential Management” on page 134
•
“Integrated Windows Authentication” on page 139
How Logins Are Used
121
How Logins Are Used
Table 10.3 Summary of How Logins Are Used
Purpose
To enable the metadata server to match an incoming user ID with a particular SAS identity
(inbound use).
Login Properties
*
User ID
To enable clients to seamlessly obtain user credentials for disparate systems (outbound use).
User ID, password, authentication domain
To designate one account as the preferred account for user access to a particular library and to make that account ID and password available to users. If you assign a login to a library, all users who can see that login use that login to access that library. This is a specialized form of outbound use that is sometimes used for a DBMS library.
User ID, password, authentication domain
To designate one host account as the account under which a particular server runs and to make that account's ID and password available to the spawner (service use).
User ID, password
*
Indicates which properties are involved. Every login should be assigned to an authentication domain.
The following figure depicts examples of login use.
Figure 10.3 Example: Use of Logins
Joe’s logins
Joe |
JoeOra | password
1
Metadata Server
OracleAuth
ETL group’s logins
ETLOra | password
2
Oracle DBMS
Oracle Libraries
SASUSERS group’s logins
ServiceOra | password
3 Special Library default login: ServiceOra
SAS General Servers group’s logins sassrv | password
4
Here are some general points about the figure:
Stored Process Server multiuser credential: sassrv
Pooled Workspace Server launch credential: sassrv
Workspace Server
(standard using SAS token authentication) launch credential: sassrv
122
Chapter 10 • Authentication Model
• The workspace server (using host authentication) is not depicted because access to that server is not usually through stored credentials.
Note: If you choose to store passwords for the workspace server, the relationships
would be comparable to the depiction of the Oracle DBMS, OracleAuth authentication domain, and Oracle logins. For example, you might put the workspace server in WorkspaceAuth and create individual and group logins in that authentication domain.
Note: A group login cannot be used to manage SAS LASR Analytic Server.
• The OLAP server is not depicted because it is not accessed using logins.
• The gray shading for the depicted workspace server indicates that this is a specialized configuration. By default the workspace server uses some form of host authentication, not SAS token authentication.
The numbers in the figure correspond to these uses:
1. Joe's first login is only for inbound use to determine his metadata identity. His password is available (cached in the user context, not stored in the metadata) but is not used to determine his identity. This login should be in DefaultAuth, but that relationship is not depicted because it is not used in determining his metadata identity.
2. Joe's second login provides seamless access to Oracle using an individual account.
This login includes a password and must be in the Oracle server's authentication domain. The ETL group's login is a shared login for the Oracle server. Joe will not use this login because his personal Oracle login has a higher priority.
3. The SASUSERS login is a designated default login for the Special Library. This login is visible to Joe (through his automatic membership in SASUSERS), so it is used when Joe accesses the Special Library. Assigning a login to a library overrides the usual login priority evaluation (which is based on identity precedence).
Note: In this example, the ServiceOra login must be in the OracleAuth
authentication domain. The list of available default logins for a library consists of only those logins that are in the associated server's authentication domain. The shading for the Special Library indicates that this is not a mainstream use. Most libraries do not have a designated login.
Note: In an alternate usage, the default library login is part of the user definition for
a service identity that provides mediated access to the library.
4. The designated launch credential for each of the depicted processing servers is stored on the SAS General Servers group definition. In this example, the servers all use the same credential. Logins that contain designated launch credentials are usually in the
DefaultAuth authentication domain, because these processing servers are usually in
DefaultAuth. However, those logins are directly paired with each server, not looked up by authentication domain. Because the authentication domain assignment for these logins is not used, the figure does not depict that assignment.
See Also
•
“How SAS Identity Is Determined” on page 114
•
“Credential Management” on page 134
•
“Identity Hierarchy” on page 17
•
“Criteria for a Designated Launch Credential” on page 126
Authentication Domains
123
Authentication Domains
What is an Authentication Domain?
An authentication domain is a name that facilitates the matching of logins with the servers for which they are valid.
Note: This matching is not important when you launch a client, but it is important when
you access certain secondary servers such as a third-party DBMS or, in some configurations, a standard workspace server.
Each user ID and password is valid within a specific scope. For example, the user ID and password that you use to log on to your computer at work are probably not the same as the user ID and password that you use to log on to a personal computer at home. It is also common for database servers and web servers to have their own authentication mechanisms, which require yet another, different, user ID and password.
An enterprise application that provides access to many different resources might require that a user have several sets of credentials. Each time a user requests access to a resource, the software must determine which credentials to use to provide access. The software could challenge the user with an interactive prompt for user ID and password, but that quickly becomes an annoyance that interrupts the user experience. The software could randomly try different credentials until it finds a set that works, but authentication attempts can be expensive in terms of performance. In the SAS Intelligence Platform, the software attempts to use only the credentials that it expects to be valid for a particular resource or system.
The software’s knowledge of which credentials are likely to be valid is based entirely on authentication domain assignments. For this reason, you must correctly assign an authentication domain to each set of resources that uses a particular authentication provider, and also assign that same authentication domain to any stored credentials that are valid for that provider.
How Many Authentication Domains Do I Need?
In the simplest case, all logins and SAS servers are associated with one authentication domain (DefaultAuth). This list describes the most common reasons for using more authentication domains:
• If you use web authentication, you might need a second authentication domain for the logins that contain web realm user IDs.
• If you have a third-party server (such as a DBMS server) that has its own user registry, you need a separate authentication domain for that server and its logins.
• If both of the following criteria are met, you need a separate authentication domain for the standard workspace server and its logins:
• The standard workspace server does not share an authentication provider with the metadata server (and cannot be configured to do so).
• You want to provide seamless individualized access to the standard workspace server.
124
Chapter 10 • Authentication Model
Outbound and Trusted Authentication Domains
Beginning in the second maintenance release for SAS 9.4 is support for outbound and trusted authentication domains. A login in an outbound domain is used only to provide
SAS applications with seamless access to external resources, such as a third-party vendor database. These logins are not part of the SAS identity phase, which attempts to determine the current metadata user by matching their authenticated user ID to the user
ID stored in a login. Therefore, for outbound domain logins, the uniqueness requirement on the user ID is not enforced.
Outbound authentication domains might also be set as trusted domains. A login in a trusted domain can be accessed only by a trusted user on behalf of the user. The identity to which the login is associated does not have direct access to the login (they do not see it listed on the Accounts tab in SAS Management Console). Instead, the login is used on their behalf by a trusted user.
Administration of logins in trusted authentication domains is performed only by unrestricted users, user administrators, and users granted the ManageCredentialMetadata permission.
You can use SAS Management Console to define and administer outbound and trusted authentication domains. See
SAS Management Console: Guide to Users and
Permissions
.
Identity Passing
About Identity Passing
When a request is passed from one system to another, it is often preferable that the requesting user identity is passed along with the request. This provides individualized access control evaluations and individual accountability in the receiving system. Here are a few examples:
• When a stored process server is launched, it has to use some SAS identity to contact the metadata server in order to discover any pre-assigned libraries.
• When a user opens a SAS folder that contains cubes, the OLAP server has to use some SAS identity to contact the metadata server to determine which cubes to show to that user.
• When a workspace server fetches SAS data from the operating system, it has to use some host identity to authenticate to the operating system and request the SAS data set files.
In each example, the identity that is used affects the outcome (for example, which libraries are available, which cubes are visible, or which data files are returned).
How SAS Servers Preserve Identity
The following table provides details about how SAS servers preserve identity.
Identity Passing
125
Table 10.4 Preservation of User Identity
Server
OLAP server
Stored process server
Workspace server (using any form of host authentication)
SAS Identity for Metadata Layer
Evaluations
Host Identity for Host Layer
Evaluations
The SAS Trusted User. However, user identity is preserved because this privileged service identity impersonates each requesting user.
*
The server's host identity. This affects drilling to detail if the underlying data is in SAS tables.
The server's host identity (the multiuser credential).
Each requesting user's SAS identity.
The exception is for library preassignment, which happens under the server identity.
**
The SAS identity of the server. This is the same as the SAS identity of the requesting user or group.
The server's host identity (the host identity of the requesting user or group).
***
Workspace server (using SAS token authentication)
Each requesting user's SAS identity.
Pooled workspace server Each requesting user's SAS identity.
The server's host identity (the launch credential).
The server's host identity (the launch credential).
Client-pooled workspace server The SAS identity of the server.
However, SAS Web Report Studio preprocesses requests based on each requesting user's SAS identity.
The server's host identity (the puddle login).
**
***
*
Except for retrieval of DBMS credentials, which happens under the SAS Trusted User's identity. This affects drilling to detail if the underlying data is in a third-party DBMS.
The requesting user is not known when library pre-assignment happens (during server initialization).
Each server process instance is launched on-demand and runs under the host identity of the requesting user (or group, if authentication is via stored group logins).
Note: You can directly specify which identity to use on a connection to the metadata
server (for example, with METAUSER and METAPASS). You can directly specify which identity to use on a connection to a host (for example, by providing explicit credentials in your code).
In the preceding table, notice that although metadata layer evaluations are almost always individualized (for example, "Does Joe have metadata layer permission to see this stored process?"), physical layer evaluations are often generalized (for example, "Does the sassrv account have host layer permissions for this data?").
In particular, requests from metadata-aware clients for SAS data often are not evaluated under the host identity of each requesting user. The host does not know the user's SAS identity, so the host cannot evaluate the user's operating system permissions. Instead, the host checks the permissions of the service account that is fetching the SAS data on behalf of the requesting user. These service accounts are referred to as launch credentials.
See Also
•
“Launch Credentials” on page 126
•
126
Chapter 10 • Authentication Model
•
Appendix 4, “Choices in Workspace Server Pooling,” on page 263
Launch Credentials
About Launch Credentials
Workspace servers and stored process servers use different process instances (SAS sessions) for different requests. Each process instance is launched by a spawner under a host account as follows:
• For the stored process server, the pooled workspace server, and a standard workspace server that uses SAS token authentication, the spawner launches process instances under a designated service account. That account is called a launch credential (or, sometimes, a multi-user credential) and is specified as part of the server's metadata definition. For example, in the initial configuration, the SAS Spawned Servers account (sassrv) is the only designated launch credential.
• For a standard workspace server that uses host authentication, the spawner launches a new process instance on-demand for each requesting user. The server's metadata definition does not include a designated launch credential. Instead, each requesting user's personal host account is used to launch that user's process instances. Or, if the server is accessed using group logins, a group account might be used. As a result, only users who can authenticate to the workspace server's host can get a workspace server.
The following figure depicts a deployment where the sassrv account is the launch credential for all servers that run under a designated account. The last server in the figure is shaded because it depicts a specialized configuration and cannot coexist within an application server that includes another standard workspace server.
Figure 10.4 Example: Launch Credentials for Processing Servers
Spawner sassrv sassrv <user> sassrv
Pooled
Workspace
Server
Stored
Process
Server
Workspace
Server
(standard)
Workspace Server
(standard using SAS token authentication)
Note: If you use client-side pooling, the puddle login is comparable to the launch
credential.
Criteria for a Designated Launch Credential
The Launch credentials (or Multiuser credentials) setting is available on the Options tab for a standard workspace server that uses SAS token authentication, the stored process server, and the pooled workspace server.
Launch Credentials
127
Figure 10.5 Choosing a Launch Credential
Note: In the preceding display, the values within parentheses indicate the authentication
domain of the credential. These values are not part of the service account ID. For example, the selected account ID is
WIN\sassrv
, not
WIN\sassrv
(DefaultAuth)
.
You should carefully select each server's launch credential for these reasons:
• Not all of the choices are viable without additional configuration.
• Each server retrieves SAS data from the operating system under the host identity of its designated launch credential. Anyone who can use a particular server can potentially access all of the data that is available to that server's launch credential.
The following list explains the choices in the launch credential list:
(None)
prevents proper functioning of the server. If the Multiuser credentials or Launch
credentials drop-down list is present and enabled on the Options tab for a server
, that server is not functional when (None) is selected.
Use spawner identity
can introduce the following complications:
• On Windows, using the spawner's host identity as a launch credential causes launched processes to run with restricted access. For example, even if the spawner runs under the local system account, the host identity of the launched process is not a member of Windows groups such as Administrators and Power
Users. The downgrade is necessary in order to avoid security exposures. The downgrade can interfere with legitimate operation of the server.
Note: In the initial configuration, a spawner on Windows runs as a service under
the local system account. You can reconfigure the spawner to run under some other account.
• In the standard configuration, the spawner's SAS identity (PUBLIC) cannot launch a pooled workspace server or a stored process server. As a PUBLIC-only identity, the spawner lacks the necessary grant of the ReadMetadata permission on the server definition.
Note: One solution is to create an individual SAS identity for the spawner
account. For example, create a user definition named SPAWNER and store the user ID of the spawner account as a login on the Accounts tab of that user definition. In the initial configuration, all registered users (SASUSERS) have
ReadMetadata permission for server definitions. If you narrow access, you might have to add a grant of ReadMetadata permission on the server's
Authorization tab for the SPAWNER user.
Note: Ensure that the server's launch credential has any necessary host access to the
SAS Web Report Studio query cache library (wrstemp) and distribution library
(wrsdist). See “Configuring SAS Web Report Studio” in SAS Intelligence
Platform: Web Application Administration Guide
.
128
Chapter 10 • Authentication Model
A listed login (for example, WIN\sassrv) or a login that you access by selecting More
logins
is the standard choice. In order to successfully function as a launch credential, a login must meet all of the following criteria:
• The login that must be visible to the spawner. All of the logins on the Accounts tab of the SAS General Servers group are visible to the spawner.
Note: The SAS Trusted User is a member of the SAS General Servers group.
The spawner uses the SAS Trusted User to gather the metadata information needed to launch servers.
• The login must include the user ID and password for an account that is known to server's host. It does not matter which authentication domain the login is in.
Note: If the server is on Windows, the user ID in the login must be in a qualified
format (for example, WIN\serviceID).
• The login must be owned by a SAS identity that has the ReadMetadata permission for the server definition. In the standard configuration, this requirement is met because the SAS General Servers group, like all other registered identities, has ReadMetadata permission for all server definitions.
However, if you choose to limit ReadMetadata permission on a server definition, you must preserve access for the SAS General Servers group.
• The login must reference an account that has host layer access to any SAS data that it retrieves.
• If the server is on Windows, the login must reference an account that has the Log
on as a batch job privilege.
Note: The login should reference a service account. This login should not
correspond to a real person.
Note: Ensure that the server's launch credential has any necessary host access to the
SAS Web Report Studio query cache and distribution libraries.
The preceding discussion assumes that your deployment uses the standard predefined metadata groups and users. You can choose to configure variations (for example, create a group other than the SAS General Servers group to hold logins for launch credentials).
In general, such variations are not recommended because they unnecessarily increase complexity and reduce consistency.
How to Create and Designate a New Launch Credential
1. Identify or create the host account.
• If the server retrieves SAS data, make sure the host account has appropriate access to those files.
• If the server is on Windows, assign the Windows privilege Log on as a batch
job to the account.
2. Log on to SAS Management Console as someone who has user administration capabilities (for example, sasadm@saspw). Other users cannot add logins to a group definition or see logins that they do not own.
3. In User Manager, create a login on the Accounts tab of the SAS General Servers group. This login must include both a user ID and a password.
Note: If the server is on Windows, the user ID in the login must be in a qualified
format (for example, WIN\serviceID).
Launch Credentials
129
Note: You can ignore any warnings about having multiple logins in the same
authentication domain. These logins are directly associated with servers. These logins are not looked up by authentication domain.
Note: You cannot reuse a login that already exists on the Accounts tab for some
other user or group. The metadata server's integrity constraints prevent you from assigning the same user ID to two different SAS identities.
Note: You can use the same login as a launch credential for more than one server.
Note: Users should not be members of the SAS General Servers group. This group
is for service identities only. Users do not retrieve these credentials, so users should not have access to these logins. In a standard configuration, these logins are available only to the privileged service identity that the spawner uses to read all server metadata (the SAS Trusted User).
4. In Server Manager, on the server's Options tab, select the login as the server's launch credential (or multiuser credential).
Note: The launch credential is designated at the level of the server component
, not at the level of the logical server .
Note: To locate a login that is not displayed in the drop-down list, select the More
logins entry. Although all logins are displayed in the secondary dialog box, any login that is not visible to the SAS Trusted User is not a viable choice.
5. To make the changes take effect, re-initialize the spawner. a. Expand the object spawner , right-click the computer , and select
Connect. b. Right-click the computer again and select Refresh Spawner. In the message box, click Yes.
c. Right-click the computer a third time and select Disconnect.
6. Validate the server.
Who Can Launch a Standard Workspace Server?
In order to launch a standard workspace server that uses any form of host authentication, a user must have access to an individual or group account that meets all of the following criteria:
• The account must be known to the workspace server's host. An internal account
(such as sasadm@saspw) does not meet this requirement.
• The account must have host-layer access to any SAS data that it retrieves.
• If the server is on Windows, and the user does not connect through Integrated
Windows authentication, the account must have the Log on as a batch job privilege.
• If the user supplies credentials interactively (for example, at a secondary logon prompt), the account must correspond to a SAS identity that has the ReadMetadata permission for the workspace server definition. In the initial configuration, all registered users meet this requirement, but a PUBLIC-only identity (someone who does not have a well-formed user definition) does not meet this requirement.
Note: If the Use Server Access Security check box on the Options tab of the
logical server is cleared, this requirement does not apply.
130
Chapter 10 • Authentication Model
See Also
•
“Identity Passing” on page 124
•
•
Appendix 4, “Choices in Workspace Server Pooling,” on page 263
•
“Windows Privileges” on page 19
PUBLIC Access and Anonymous Access
About PUBLIC Access and Anonymous Access
In general, only users who can authenticate and who have a well-formed user definition should use a SAS deployment. However, in order to accommodate scenarios where more general access is desired, the following specialized configurations are supported:
• PUBLIC access enables unregistered users to participate if they can authenticate to the metadata server (directly or through a trust mechanism). Unregistered users are referred to as PUBLIC-only users because their only SAS identity is that of the
PUBLIC group. A PUBLIC-only user has the logins, permissions, and capabilities of the PUBLIC group. A PUBLIC-only user cannot belong to any other groups, or have any personal logins, or have any specialized (individual) access controls. Not all applications allow a PUBLIC-only user to log on.
• Anonymous access enables unregistered users to participate without authenticating to the SAS environment. Anonymous access is an optional configuration that is available for only a few applications. Anonymous access is supported only with SAS authentication; anonymous access is not compatible with web authentication.
Anonymous access is supported as follows:
• For SAS BI Web Services and the SAS Stored Process Web Application, a user who connects through anonymous access uses the SAS Anonymous Web User identity. This is a service identity that functions as a surrogate for users who connect without supplying credentials. See “Using the SAS Anonymous Web
User with SAS Authentication” in SAS Intelligence Platform: Middle-Tier
Administration Guide
.
• For the SAS Information Delivery Portal, a user who connects through anonymous access uses the Unchallenged Access User identity. This is a service identity that functions as a surrogate for users who connect without supplying credentials. See “Enabling Unchallenged Portal Access” in SAS Intelligence
Platform: Web Application Administration Guide
.
PUBLIC access and anonymous access differ in the following ways:
• In PUBLIC access, each participating user must authenticate. In anonymous access, participating connections do not require user authentication.
• In PUBLIC access, participating users share the PUBLIC group identity. In anonymous access, participating connections share a designated service identity (the surrogate identity is always a member of both the SASUSERS group and the
PUBLIC group).
• You can choose to provide wide support for PUBLIC access. You cannot extend support for anonymous access beyond the specific applications that can be configured to use it.
PUBLIC Access and Anonymous Access
131
CAUTION:
If you choose to offer PUBLIC or anonymous access, you risk users seeing more data and content than you might expect.
Carefully review and manage access control for the PUBLIC group. If you offer anonymous access, carefully review and manage access control for your surrogate service identity too.
How to Enable PUBLIC Access
You can use the following instructions to set up a specialized configuration in which unregistered users can participate from most SAS applications. However, not all SAS applications can be configured to accept PUBLIC-only users. For example, SAS Visual
Analytics never accepts PUBLIC-only users.
To enable PUBLIC access, follow these steps in SAS Management Console:
1. Provide the necessary repository-level access.
On the Plug-ins tab, under Authorization Manager
ð
Access Control Templates, right-click the repository ACT (Default ACT) and select Properties. On the
Permission Pattern tab, grant the ReadMetadata and WriteMetadata permissions to
PUBLIC.
Note: Even users who consume only content need both of these permissions at the
repository level, because some applications write system information about user activity, even during what appears to be a view-only transaction.
2. Provide Read access as needed.
On the Folders tab, give the PUBLIC group the Read permission for any information maps, cubes, and MLE data that you want to make universally available.
A good approach is to create a folder branch for such content, set the grant on the top folder in that branch, and allow the grant to flow through the branch.
Note: Users also need ReadMetadata permission to folders and content objects. In
general, it is not necessary to set specific grants because this permission must flow through from the repository ACT into the public areas of the folder tree (for navigational purposes).
Note: If you want to allow everyone (including unregistered users) to contribute
content to a particular folder, give the PUBLIC group a grant of the
WriteMemberMetadata permission on that folder's Authorization tab.
3. Review role assignments for the PUBLIC group.
On the Plug-ins tab, under User Manager, right-click the PUBLIC group and select
Properties. Review the PUBLIC group's role memberships. Often, no adjustments are necessary, because the initial role assignments give the PUBLIC group basic capabilities.
4. Make sure that the PUBLIC group can use servers. a. On the Plug-ins tab, under Server Manager, verify that the PUBLIC group has the ReadMetadata permission for any servers that the PUBLIC-only users access.
b. If necessary, add one or more logins on the PUBLIC group's Accounts tab (for example, to provide seamless access to a third-party DBMS).
c. If you have configured client-side pooling, verify that PUBLIC is a designated puddle group.
5. Configure middle tier properties applications to accept PUBLIC-only users.
132
Chapter 10 • Authentication Model
On the Plug-ins tab, navigate to Application Management
ð
Configuration
Manager, and make changes as needed. Not all deployments include and use all components.
T I P
If you need additional information, see the administrative documentation for your application. Not all deployments include and use all components. Not all middle-tier components have an applicable property.
a. For SAS Web Report Studio, set the Allow Public Users property to Yes (this property is in the Application Configuration section of the Settings tab of the properties dialog box for Web Report Studio).
b. For the SAS Stored Process Web application, set the App.PublicIdAllowed property to true (this property is on the Advanced tab of the properties dialog box for the Stored Process Web App).
c. To make the changes take effect, restart the SAS Web Infrastructure Services
Application and then restart the affected web applications.
133
Chapter 11
Authentication Mechanisms
Summary of Methods for LDAP Integration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Introduction to Authentication Mechanisms
This chapter describes the following features:
• Internal authentication mechanisms unify the SAS realm and provide a degree of independence from your general computing environment. The internal mechanisms are SAS internal authentication and SAS token authentication.
• External authentication mechanisms integrate SAS into your computing environment. External mechanisms include direct LDAP authentication, host authentication (credential-based), Integrated Windows authentication, and web authentication.
• Credential management provides single sign-on through reuse of cached credentials or retrieval of stored passwords.
• Pluggable authentication modules (PAM) extend UNIX host authentication.
134
Chapter 11 • Authentication Mechanisms
• Trust relationships facilitate communication to the metadata server by permitting one privileged account to connect on behalf of other users (trusted user) or by accepting requests that use a proprietary protocol (trusted peer).
Credential Management
Table 11.1 Credential Management
Summary A supporting feature in which clients reuse cached credentials or retrieve stored credentials. Clients use authentication domain assignments to determine which credentials are valid for which servers. The target server validates the client-supplied credentials against its authentication provider.
Scope From clients that are already connected to the metadata server to third-party servers, the Scalable
Performance Data Server, and, in some cases, the workspace server.
Provides access to servers using individual or shared accounts.
Benefits
Limits • Involves passing user IDs and passwords across the network.
• Can involve maintaining SAS copies of external passwords.
Use Always available.
Credential management techniques populate an in-memory list of credentials for each connected user. Each list is called a user context and includes these entries:
• If the user interactively provides credentials when launching a SAS client, those credentials are added to the list, with these exceptions:
• If web authentication is configured, the password from a user's interactive logon to a web client is not available to be added to the list.
• If the user logs on with Integrated Windows authentication (IWA), the user's password is not available to be added to the list.
• If the user interactively provides credentials at any point in the session, those credentials are added to the list.
• If the user's metadata definition has logins that include passwords, those credentials are added to the list.
• If the user belongs to any groups whose metadata definition has logins that include passwords, those credentials are added to the list.
Note: Credentials from a user or group's metadata definition are not included in the
initial list that is created when a user logs on. Instead, such credentials are added to the list dynamically (when and if they are needed in the course of the user's session).
The following table depicts an example of the contents of a user context:
Table 11.2 Example: Contents of a User Context
User ID
myWinID
Password
********
Authentication Domain
DefaultAuth
User ID
GroupDBMSid
Password
********
Credential Management
135
Authentication Domain
DBMSauth
Notice that each entry is assigned to an authentication domain. This enables pairing of credentials with the servers for which they are valid. The entries are created as follows:
• The client creates the first entry by caching and inserting the user ID and password that a user submits in an initial logon. This makes the user's DefaultAuth password available for reuse even though that password is not stored in the metadata. The client automatically assigns the first entry to the DefaultAuth authentication domain except in these circumstances:
• The user's connection profile contains a user ID with an @saspw suffix.
Note: Notice that this circumstance describes the user ID in the user's connection
profile, not the user ID that the user supplies in the logon dialog box.
• The Authentication domain field in the user's connection profile contains a value other than DefaultAuth (and is not empty).
• The user is accessing a web client at a site that has configured web authentication
(or has specified a different authentication domain in the web configuration for some other reason).
• The client creates the second entry by retrieving information from the metadata (in the preceding example, from a group that the user belongs to). Such logins are for outbound use, so they must include a password and an appropriate authentication domain assignment.
When a user requests access to a server that requires credential-based authentication, the client completes these steps:
1. Examine the server's metadata to determine which authentication domain the server belongs to.
Note: In SAS Management Console, this information is displayed on the Options
tab of each of the server's connection objects .
2. Examine the user's context to determine whether it includes any credentials that are assigned to the target server's authentication domain. The process is as follows:
• If the context includes a cached entry for the target authentication domain, that entry is used.
• If the user context contains a retrieved entry for that authentication domain, that entry is used. If there is more than one retrieved entry for an authentication domain, the entry that is closest to the user is used.
• If there is an identity precedence tie among retrieved entries (for example, if a user is a direct member of two groups and both groups have logins in the relevant authentication domain), the same login is used consistently, but you cannot control which of the two logins is used.
• If the user context contains no entries in the target authentication domain, desktop clients prompts the user for credentials. Web applications cannot prompt.
Note: SAS Web Report Studio has an interactive password management feature.
3. Present the credentials to the target server for authentication against its provider.
Here are some additional tips:
136
Chapter 11 • Authentication Mechanisms
• Because the credentials that are added to a user context from an initial logon are not likely to be valid for a third-party DBMS, you usually have to store credentials for such servers.
• Authentication domains have no effect on an initial logon. Authentication domains affect access to secondary servers that perform credential-based authentication.
• To prevent attempts to reuse internal credentials for the workspace server (which does not accept internal credentials), users who have an internal account should use a dedicated connection profile for that account. In their internal connection profiles, users should leave the Authentication domain field blank (or specify an arbitrary value such as InternalAuth).
See Also
•
“How Logins Are Used” on page 121
•
“Identity Hierarchy” on page 17
•
“How to Store Passwords for a Third-Party Server” on page 175
Direct LDAP Authentication
Table 11.3 Direct Use of LDAP Authentication
Summary The metadata server validates some users against an LDAP provider such as Active Directory. Direct
LDAP enables the metadata server to recognize accounts that are not known to its host; direct LDAP does not modify the host's behavior.
Scope • Primarily used for connections to the metadata server.
• Can also be used for direct connections from a data provider to the OLAP server.
Benefits Enables users to use their Windows accounts to authenticate to a metadata server that runs on UNIX.
Limits
Use
• Not an alternative to storing user IDs in the metadata (that requirement applies to all configurations).
• Not supported for workspace servers or stored process servers.
• Might involve appending a special qualifier to user IDs that are stored in the metadata.
Optional.
The following figure contrasts back-end use and direct use.
See Also
Host Authentication
137
Figure 11.1 Comparison of Back-End and Direct Use of LDAP
Back-End Use
LDAP or
Active Directory authentication
3
ID & password
2
ID & password
Host
4
ID
1
ID & password
8 accept
Metadata
Server
6
ID
5
ID
Client
7
SAS identity
Direct Use
2
ID & password
LDAP or
Active Directory authentication
Metadata
Server
3
ID
4
ID
5
SAS identity
Configuring the metadata server to directly use LDAP is one of several methods for
.
•
“How to Configure Direct LDAP Authentication” on page 160
•
“How to Configure TLS between the Metadata Server and an LDAP Server” on page
•
“Pluggable Authentication Modules (PAM)” on page 142
Host Authentication
Table 11.4 Host Authentication (Credential-Based)
Summary A client supplies an external user ID and password to a SAS server. The SAS server passes the credentials to its host for authentication.
Note: Another form of host authentication, Integrated Windows authentication (IWA), is documented separately.
Scope • Primarily used for direct connections to the metadata server or OLAP server.
• Not used for metadata-aware connections to the OLAP server or stored process server.
• Sometimes used for connections to the workspace server.
Benefits No configuration is required. Can enable users to log on to SAS applications with the same credentials that they use in your general computing environment.
138
Chapter 11 • Authentication Mechanisms
Limits
Use
• On a workspace server on Windows, requires that users have the Windows privilege Log on as a batch
job.
• Involves passing user IDs and passwords across the network.
Always available.
The following figure shows one example of how this mechanism works:
Figure 11.2 Host Authentication (credential-based)
2
3
ID & password
Authentication by Active Directory,
Local Security Authority, PAM,
UNIX password file structure, or other provider.
ID & password
Host
4
ID
5
ID
1
ID & password
6 accept
Target
Server
Client
The numbers in the figure correspond to these actions:
1. The client obtains the user's ID and password (interactively or through credential management). The client sends those credentials to the target server.
2. The server passes the credentials to its host for authentication.
3. The host passes the credentials to its authentication provider.
4. After verifying that the user ID and password correspond to a valid account, the host's authentication provider returns the user's ID to the host.
5. The host returns the user's ID to the SAS server.
6. The server accepts the client connection.
See Also
•
“Windows Privileges” on page 19
•
“About the Workspace Server's Options Tab” on page 179
•
“Integrated Windows Authentication” on page 139
•
“Authentication to the Metadata Server” on page 114
•
“Summary of Methods for LDAP Integration” on page 151
Integrated Windows Authentication
139
Integrated Windows Authentication
Table 11.5 Integrated Windows Authentication (IWA)
Summary Participating SAS servers accept users who have successfully authenticated to their Windows desktop.
Scope For Windows desktop clients and SAS servers on Windows and UNIX.
• Primarily used for connections to the metadata server and the standard workspace server.
• Also supported for direct connections to an OLAP server (for example, from a data provider).
Benefits • Bypasses the initial logon prompt.
• Accommodates logon mechanisms that are not password-based (such as smart cards or biometrics).
• No user credentials are transmitted.
• Users do not need the Log on as a batch job privilege. See “Windows Privileges” on page 19
.
140
Chapter 11 • Authentication Mechanisms
Limits
Use
• All participating clients and servers must authenticate against the same Windows domain (or against domains that trust one another).
• If you use IWA for the metadata server, there are no cached credentials from an initial logon. For this reason, it is a good idea to configure IWA for the standard workspace server also, if possible (IWA is not supported on z/OS).
• Web applications cannot participate in this implementation of IWA. However, if you configure web authentication, and your web environment offers IWA, then your web applications can use IWA. To configure IWA for the SAS Web Application Server, see “Configuring SAS Web Application Server” in
SAS Intelligence Platform: Middle-Tier Administration Guide
.
• Desktop clients that run on UNIX (for example, SAS Management Console on UNIX) cannot participate in IWA.
• If your SAS servers use DNS aliases, you must manually register those aliases in order to support
Kerberos based IWA connections. See “Registering SPNs” on page 169
.
• Additional limits for servers on Windows:
• If you use IWA for a Windows workspace server that accesses Windows network resources (for example, remote data access), the Kerberos protocol must be used and the object spawner account
must have the trusted for delegation privilege. See “Windows Privileges” on page 19
.
• Additional limits for servers on UNIX:
• In order to use IWA on UNIX platforms:
• For SAS 9.4 and the first maintenance release for SAS 9.4 on all platforms, you must purchase, install, and configure an additional third-party product (Quest Authentication Services 4.0).
• For the second maintenance release for SAS 9.4 on Linux platforms, you must ensure that a shared library that implements the GSSAPI with Kerberos 5 extensions is installed and configured to allow authentication against your Active Directory domain or Kerberos realm. Quest
Authentication Services fulfills this requirement, as do the krb5 packages provided in supported operating system distributions and in various third-party solutions.
• For the second maintenance release for SAS 9.4 on all other UNIX platforms, you must purchase, install, and configure an additional third-party product (Quest Authentication Services 4.0).
• When you use IWA on UNIX, only Kerberos connections are supported (there is no support for
NTLM on UNIX).
• If you use IWA for a UNIX workspace server that makes outbound Kerberos requests, the service principal account in Active Directory must have the trusted for delegation to all services privilege.
See “Windows Privileges” on page 19
.
Optional.
The following figure is an abstraction of how this mechanism works.
See Also
Integrated Windows Authentication
141
Figure 11.3 Integrated Windows Authentication
Windows authentication, trusted authority
1 request for an authenticated token
2 authenticated token
4 authenticated token
3 authenticated token
6 accept
Desktop Client on Windows
5 accept
Target Server
The numbers in the figure correspond to these actions:
1. The client asks Windows for a token that represents the user who is currently logged on to the client computer. This step is initiated when a user launches a desktop client or makes a request for a workspace server.
Note: The token represents the client who is running the SAS application. If you use
the Windows
runas
command to launch a SAS application, the token represents the host account that you chose to use.
2. Windows provides the token to the client.
3. The client sends the Windows token to the target server. Notice that only the token is sent; the user's password is not available to the target server.
4. The target server sends the token back to Windows for verification.
5. Windows tells the target server that the token is valid.
6. The target server accepts the connection from the client. By default, the user's authenticated ID is returned to the target server as follows:
• If the target server is on Windows, the authenticated user ID is returned in domain qualified format (for example, WIN\joe).
• If the target server is on UNIX, the authentication user ID is returned in short format; the user ID is not qualified (for example, joe).
Note: The format in which the authenticated user ID is returned to the target server
must match the format in which that user ID is stored in the SAS metadata. See
•
“How SAS Identity Is Determined” on page 114
•
“How to Configure Integrated Windows Authentication” on page 165
142
Chapter 11 • Authentication Mechanisms
Pluggable Authentication Modules (PAM)
Table 11.6 PAM (Pluggable Authentication Modules)
Summary A supporting feature that extends UNIX host authentication to recognize an additional provider such as
Active Directory. When a SAS server asks its UNIX host to validate a user's credentials, the host sends the user's ID and password to the configured additional provider for verification.
PAM extends the host's authentication process to recognize an additional provider; PAM does not modify the metadata server's behavior.
Scope
Benefits
Limits
Use
Affects all SAS servers that run on the UNIX host and rely on the host operating system to authenticate users. Typically, the metadata server and the workspace server use host authentication.
Can be used to enable users to use their Windows accounts to authenticate to a metadata server or workspace server that run on UNIX.
Not an alternative to storing user IDs in the metadata (that requirement applies to all configurations).
Optional.
This mechanism is useful if both the metadata server and the workspace server are on
UNIX and you want users to use Windows accounts to access these servers.
This mechanism can also be useful if one of these servers is on Windows, the other is on
UNIX, and you want to avoid credential prompts for the workspace server. However, if you use PAM to resolve a mixed provider situation, users who access the workspace server must have two logins. One login should include the user's ID in its qualified form.
The other login should include the same ID in short (unqualified) form. Both logins should be in the DefaultAuth authentication domain. Neither login should include a password. For example, a user's logins might look like this:
DefaultAuth | WIN\joe | (no password)
DefaultAuth | joe | (no password)
For configuration instructions, see Configuration Guide for SAS Foundation for UNIX
Environments at
http://support.sas.com/documentation/installcenter .
See Also
•
“Summary of Methods for LDAP Integration” on page 151
•
“Direct LDAP Authentication” on page 136
•
SAS Internal Authentication
143
SAS Internal Authentication
Table 11.7 Internal Authentication
Summary Participating SAS servers validate incoming user IDs that have a special suffix (@saspw) against a list of accounts that exist only in the metadata.
Scope
Benefits
• Primarily used for connections to the metadata server.
• Also supported for direct connections to the OLAP server. This is not a common use. In the SQL procedure, a connection to the OLAP server can be made using an internal account.
• Minimizes the need to create external accounts for service identities.
• Facilitates intermittent use of an administrative role by enabling a user to have a second metadata identity without having a second external account.
• Provides independence from the rest of your computing environment. For example, internal accounts do not have to follow your general password change requirements, cannot be used to access resources beyond the SAS metadata, and are not affected by changes in machine names or in your authentication configurations.
Limits
Use
• Internal accounts are not intended for regular users (they are intended for only metadata administrators and some service identities).
• Someone who has only an internal account cannot seamlessly access the workspace server.
• An internal account cannot participate in Integrated Windows authentication or web authentication.
• You cannot use an internal account to delete, unregister, add, or initialize a foundation repository.
• You cannot run a server under an internal account. For example, the SAS General Server User (sassrv) must be a host account.
Always available.
Internal accounts exist only in the metadata and can be created in SAS Management
Console. The following displays depict the General tab and Accounts tab for an administrator named Joe. Notice that Joe's internal user ID is Joe@saspw. The user ID for an internal account is always in the format name
@saspw
. The name comes from the
Name field on the user's General tab.
The following figure depicts the internal authentication process:
144
Chapter 11 • Authentication Mechanisms
Figure 11.4 Internal Authentication
1 joe@saspw
& password
2 joe@saspw
& password
4 accept
Client
Metadata
Server
login login login login login login
account
Joe
login
account sastrust account sasadm
3
Joe
Metadata Repository
The numbers in the preceding figure correspond to these actions:
1. At a logon prompt, Joe enters his internal credentials. The client sends those credentials to the metadata server for verification.
2. The metadata server recognizes that the ID is for an internal account (because the ID has the @saspw suffix), so the metadata server checks the credentials against its list of internal accounts.
3. After validating the ID and password, the metadata server accepts the client connection. The connection is under the identity of the SAS user who owns the account.
Note: Joe can have logins in addition to his internal account, but he does not need a
login to establish his SAS identity.
Here are some tips for working with internal accounts:
• There are two distinct expiration settings. Do not confuse the account expiration date with the password expiration period.
• If repeated attempts to log on with an internal account fail, make sure you are including the @saspw suffix in the user ID. Another cause of failure is the account is locked.
• If you have both an internal account and an external account, use a dedicated connection profile for your internal account. In that profile, leave the Authentication
domain field blank. This ensures optimal credential reuse.
• If you do not have user administration capabilities, you cannot see internal accounts, unless you are viewing your own definition and you happen to have an internal account.
Because internal accounts exist only in the metadata, they do not automatically follow the policies of other authentication providers. Here are the initial policies for internal accounts:
• Accounts do not expire and are not suspended due to inactivity.
• Passwords must be at least six characters, do not have to include numbers or mixed case, and do not expire.
• The five most recent passwords cannot be reused.
See Also
SAS Token Authentication
145
• After three failed attempts to log on, an account is locked for one hour. An administrator can unlock the account by accessing the Accounts tab in the user's definition in SAS Management Console.
• A forced password change occurs on first use and after a password is reset. This policy applies only to accounts with passwords that periodically expire. By initial policy, passwords do not expire, so forced password changes do not occur.
You can set all of these policies globally (at the server-level). You can also selectively override many of these policies on a per-account basis.
•
“Authentication to the Metadata Server” on page 114
•
“How to Change Internal Account Policies” on page 177
•
“Unlock an Internal Account” on page 37
SAS Token Authentication
Table 11.8 SAS Token Authentication
Summary The metadata server generates and validates a single-use identity token for each authentication event. This has the effect of causing participating SAS servers to accept users who are connected to the metadata server.
Scope
Benefits
Limits
Use
• Primarily used for metadata-aware connections to the stored process server, the server-side pooled workspace server, the OLAP server, the content server, and (in a specialized configuration) the standard workspace server.
• Also used by launched servers to connect back to the metadata server (for example, from the workspace server to the metadata server for library pre-assignment).
• Preserves client identity for metadata layer access control and auditing purposes.
• No individual external accounts are required, no user passwords are stored in the metadata, and no reusable credentials are transmitted.
• On the workspace server, reduces granularity of host access.
• Supported only for metadata-aware connections (in which the client learns about the target server by reading the server's metadata definition).
Optional for the workspace server, otherwise mandatory within its scope.
The following figure is an abstraction of how this mechanism works.
146
Chapter 11 • Authentication Mechanisms
Figure 11.5 SAS Token Authentication
Client that already has a connection to the metadata server
Metadata Server
(generates identity tokens)
1 request for an identity token
2 identity token
4 identity token
3 identity token
6 accept
5 accept
Target
Server
The numbers in the figure correspond to these actions:
1. Over the user's existing connection to the metadata server, the client requests an identity token for the target server. This step is initiated by a user request that requires access to the target server (for example, by a request in SAS Enterprise
Guide for a cube that is associated with the OLAP server).
2. The metadata server generates the token and sends it to the client.
3. The client provides the token to the target server.
4. The target server sends the token to the metadata server for validation.
5. The metadata server validates the token and returns an acceptance message and a representation of the user to the target server.
6. The target server accepts the connection.
See Also
•
“How to Configure SAS Token Authentication” on page 157
•
Trusted Peer Connections
Table 11.9 Trusted Peer
Summary The metadata server accepts peer SAS sessions and servers that connect using a proprietary protocol
(trusting that those connecting identities have already been properly authenticated).
From any SAS session or SAS IOM server process to the metadata server. The scope is configurable.
Scope
Benefits • Enables SAS/CONNECT servers to access the metadata server.
• Can enable batch processes to connect to the metadata server.
• In a clustered metadata server environment, can support communication among nodes within the cluster.
For information about clustered metadata servers, see
SAS Intelligence Platform: System Administration
Guide
Limits
Use
Trusted Peer Connections
147
• On Windows, it is safer to use Integrated Windows authentication (IWA). By default, a clustered metadata server environment on Windows uses IWA for connections among nodes.
• For greater security, minimize availability of this feature by modifying the trustedPeers.xml file, as documented in this topic.
• For a clustered metadata server on UNIX or z/OS, trusted peer connections by the installer ID and from the cluster machines must be allowed.
• If you use your operating system scheduler to run metadata backup jobs, trusted peer connections from the host account that runs those jobs must be allowed.
If the metadata server's start command includes the TRUSTSASPEER= option, the referenced trustedPeers.xml file specifies which user IDs and machines are eligible to connect to the metadata server using the trusted peer protocol.
By default, all user IDs and machines are eligible. The initial contents of the trustedPeers.xml file are as follows:
<TrustedSASPeers>
<TrustedSASPeerClients>
<client name="SAS" />
</TrustedSASPeerClients>
<TrustedSASPeerUsers>
<user name="*" />
</TrustedSASPeerUsers>
<TrustedSASPeerMachines>
<machine ip="*" />
</TrustedSASPeerMachines>
</TrustedSASPeers>
For greater security, we recommend that you target this mechanism so that the metadata server does not accept every connection that uses the proprietary protocol. You can use either or both of these constraints:
• accept only specified user IDs
• accept only connections that originate from specified machines
You can define constraints in trustedPeers.xml as follows:
TrustedSASPeerClients lists eligible client types.
SAS
and
java
are the valid values. It is recommended that you reject connections from Java clients.
Typically, there is only one entry between the
TrustedSASPeerClients
tags:
<client name="SAS"/>
TrustedSASPeerUsers lists eligible user IDs. To represent all users, use an asterisk (*). To represent all users in a Windows domain, use the format *@domain. For Windows accounts, provide domain-qualified (or machine-qualified) IDs. For example, you might insert these three entries between the
TrustedSASPeerUsers
tags:
<user name="*@winXP"/>
148
Chapter 11 • Authentication Mechanisms
<user name="tara"/>
<user name="batchjobID"/>
TrustedSASPeerMachines lists eligible points of origin. Identify machines by IP address. You can use asterisks
(*) as wildcards. For example, you might insert these three entries between the
TrustedSASPeerMachines
tags:
<machine ip="1.2.3.4"/>
<machine ip="A:B:C:D:E:F:1.2.3.4"/>
<machine ip="*.*.8.9"/>
Note: An additional constraint, TrustedSASDomains, is supported for backward
compatibility but will be deprecated in a future release.
If you modify the trustedPeers.xml file, keep the following points in mind:
• Only connections that meet all specified criteria are accepted. If any of the sections are empty, no trusted peer connections are allowed.
• For a clustered metadata server on UNIX or z/OS, you must preserve support for trusted peer connections from the cluster nodes (under the install account) as follows:
• In the TrustedSASPeerUsers section of the trustedPeers.xml file, include the installer account (or accounts).
• In the TrustedSASPeerMachines section of the trustedPeers.xml file, include the
IP address of each cluster machine.
Note: The trustedPeers.xml file is in your equivalent of
SAS/Config/Lev1/
SASMeta/MetadataServer/
.
Trusted User Connections
Table 11.10 Trusted User
Summary The metadata server allows a privileged account to act on behalf of other users (trusting that those users have already been properly authenticated).
Scope
Benefits
To the metadata server from the object spawner, the OLAP server, SAS web applications (if web authentication is used), and batch report processes.
Supports the optional web authentication configuration. Enables the OLAP server and the object spawner to impersonate each requesting user on connections to the metadata server. Enables batch reporting processes to connect to the metadata server under their identities.
Limits
Use
It is important to protect this privileged account.
Required.
In a new deployment, the trustedUsers.txt file lists one account that serves as the trusted user for the entire deployment.
CAUTION:
Web Authentication
149
Do not add regular users to the trustedUsers.txt file.
The trusted user is a privileged service identity that can act on behalf of all other users.
Note: The trustedUsers.txt file is in your equivalent of
SAS/Config/Lev1/
SASMeta/MetadataServer/
.
Web Authentication
Table 11.11 Web Authentication
Summary The metadata server accepts users who are authenticated at the web perimeter.
Scope
Benefits
From SAS web applications to the metadata server.
• Facilitates single sign-on from and across the web realm.
• Enables SAS web applications to use whatever authentication scheme you have set up in your web environment.
• Can reduce the number of user accounts that you have to create in the metadata server's authentication provider, because users who use only web applications no longer need accounts with the metadata server's provider.
Limits
Use
• When you use web authentication to access the metadata server, there are no cached credentials from an initial logon. This prevents authentication from web applications to the standard workspace server through reuse of credentials.
• Prevents users from logging on to a SAS web application with a SAS internal account.
• Not compatible with anonymous access. See “About PUBLIC Access and Anonymous Access” on page
.
Optional.
The following figure depicts the high-level choice in authentication method for SAS web applications:
• SAS authentication (any form of authentication in which the metadata server is responsible for requesting verification)
• web authentication (any form of authentication in which verification occurs in the web realm and the metadata server trusts that verification)
150
Chapter 11 • Authentication Mechanisms
Figure 11.6 Examples of SAS Authentication and Web Authentication
SAS Authentication
1 request
2 prompt
3 credentials
Web
Application
Server
4 credentials
7
Metadata
Server
5
Authentication
Provider
6
Metadata
Repository
Web Authentication
1 request
2 prompt
3 credentials
Web
Application
Server
4
5
ID
trusted conne ction
7
Metadata
Server
6
Authentication
Provider
Metadata
Repository
The preceding figures are simplified in order to highlight the differences between the two configurations. The following figure includes additional detail about web authentication.
Figure 11.7 Web Authentication: A Closer Look
1 request
2 prompt
3 credentials
Web Application Server
SAS Web Report Studio
Trusted Login Module
Local User Service
6 7
SAS Logon Manager
5
Realm
4
Authentication
Provider
8
ID
truste d connec tion
Metadata
Server
9
Metadata
Repository
Summary of Methods for LDAP Integration
151
In the preceding figure, a user who is not already authenticated at the web perimeter makes a request to access SAS Web Report Studio. The numbers in the figure correspond to the following activities:
1. In a web browser, the user makes a request for the target application (in this example, the target application is SAS Web Report Studio).
2. The web container prompts the user for credentials.
Note: If the user has already authenticated at the web perimeter, this step is omitted.
3. The user supplies credentials to the web container.
Note: If the user has already authenticated at the web perimeter, this step is omitted.
4. The web container's realm directs the container to authenticate the user against a designated third-party authentication provider.
5. SAS Logon Manager retrieves the authenticated user ID from the web container. The
SAS Logon Manager stores the user ID for later use.
6. SAS Logon Manager provides a ticket for the target application and redirects the user to that application.
7. The target application connects back to SAS Logon Manager (over HTTP) to validate the supplied ticket. SAS Logon Manager supplies the requesting user’s authenticated user ID as part of the validation response.
8. The target application uses the Trusted Login Module (from the SASTRUSTED
JAAS context) to generate a JAAS subject from the authenticated user ID. The generated JAAS subject is given to the local user service instance (inside the target application), which creates a user context and connects to the metadata server.
9. The metadata server looks up the user's ID in the metadata repository in order to determine the user’s SAS identity. As usual, this step does not involve password validation and is not affected by authentication domain assignments. Only the user's
ID is being matched (the authentication domain assignment in a login affects only credential reuse).
Note: Some components (such as SAS Comment Manager) do not require a connection
to the metadata server. Such components do not have a user context or a persistent connection to the metadata server.
See Also
“How to Configure Web Authentication” on page 158
Summary of Methods for LDAP Integration
SAS supports the following methods for integration with LDAP: host use of LDAP
The SAS server’s host uses an LDAP provider as a back-end authentication provider.
From the perspective of the SAS server, this is host authentication. For example:
• Active Directory is the standard back-end authentication provider on Windows.
• Some UNIX hosts recognize LDAP accounts (or can be configured to do so). See
“Pluggable Authentication Modules (PAM)” on page 142 .
152
Chapter 11 • Authentication Mechanisms
sasauth use of LDAP (UNIX only)
This method provides a direct connection from sasauth (the UNIX host authentication module) to an LDAP database for authentication. This method provides an authenticated UNIX host identity for each user. For configuration instructions, see Configuration Guide for SAS Foundation for UNIX Environments at http://support.sas.com/documentation/installcenter .
metadata server use of LDAP
The metadata server validates some users against an LDAP provider such as Active
Directory. This method enables the metadata server to recognize accounts that are not known to its host. It does not provide SAS with an authenticated UNIX host
identity for each user. See “Direct LDAP Authentication” on page 136 .
LDAP integration support is for authentication purposes only, not for authorization.
Summary for Single Sign-On
There is no individual mechanism that provides end-to-end single sign-on (SSO). The following authentication processes are transparent:
• Integrated Windows authentication (IWA) is based on previous authentication to your desktop and provides silent launch for SAS desktop applications (and, sometimes, silent access to the workspace server).
• Web authentication is based on previous authentication to your web realm and provides silent launch for SAS web applications.
• SAS token authentication requires a connection to the metadata server and provides silent access to most SAS servers.
• Credential reuse and retrieval requires a connection to the metadata server and can provide silent access to any server.
Some configurations can interfere with SSO to back-end servers. This table summarizes the considerations:
Table 11.12 SSO Considerations for Selected Authentication Mechanisms
Feature
Internal authentication
SAS token authentication
IWA
Front-end
SSO
Back-end
SSO Notes
An internal account cannot participate in IWA or web authentication.
Facilitates SSO to most SAS servers.
Web authentication
Direct LDAP authentication
Facilitates silent launch of desktop applications. If not fully configured, prevents SSO to a standard workspace server.
*
Facilitates silent launch of web applications. Prevents SSO to a standard workspace server.
*
Not compatible with silent launch. Prevents SSO to a standard workspace server.
*
Summary by Server Type
153
Feature
PAM
Front-end
SSO
Back-end
SSO Notes
Can help unify authentication.
Credential
Management
Facilitates SSO to third-party servers and (in some configurations) workspace servers.
*
Unless the server is configured for SAS token authentication or accessed using stored credentials.
Summary by Server Type
•
•
•
•
This table provides a high-level review of support for different authentication mechanisms. In the table, the following symbols indicate the extent to which each server can be accessed using each feature:
Supported
Supported for only direct connections (not metadata aware connections)
Intended for only administrators and some service identities
Used only for certain server-to-server communications
• Not supported
Table 11.13 Summary: How Servers Can Be Accessed
Server Type
OLAP
Stored Process or
Pooled Workspace
Client-
Pooled
Workspace
*
Mechanism
Host authentication
(credentials)
Integrated Windows authentication
Web authentication
Direct LDAP authentication
Internal authentication
SAS token authentication
Trusted user
Trusted peer
Metadata Workspace
*
For the client-pooled workspace server, user access depends on membership in a puddle group.
154
Chapter 11 • Authentication Mechanisms
155
Chapter 12
Authentication Tasks
How to Configure SAS Token Authentication . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Step-By-Step Instructions for Web Authentication . . . . . . . . . . . . . . . . . . . . . . . . 159
Logins for Users Who Participate in Web Authentication . . . . . . . . . . . . . . . . . . . 159
How to Configure Direct LDAP Authentication . . . . . . . . . . . . . . . . . . . . . . . . . . .
How to Configure Integrated Windows Authentication . . . . . . . . . . . . . . . . . . . . .
How to Store Passwords for the Workspace Server . . . . . . . . . . . . . . . . . . . . . . . .
How to Store Passwords for a Third-Party Server . . . . . . . . . . . . . . . . . . . . . . . . .
About the Workspace Server's Options Tab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
156
Chapter 12 • Authentication Tasks
How to Facilitate Authentication
Identify or Create User Accounts
Each user must have an account that provides access to the metadata server, either directly or through a trust relationship. Determine which of the following situations applies to you and complete any tasks as indicated.
• In the simplest case, users already have accounts that are known to the metadata server's host, so no action on your part is required. For example, the metadata server is on UNIX, and users have accounts in an LDAP provider that the UNIX host recognizes. Or the metadata server is on Windows, and users have Active Directory accounts.
• In some cases, users have accounts that are not currently recognized by the metadata server's host. Consider the examples in the following table.
Table 12.1 Incorporating Unrelated Accounts
Scenario
You have Active Directory accounts but the metadata server is on UNIX.
Possible Solution
Enable the UNIX host to recognize the
accounts. See “Pluggable Authentication
.
You have accounts in an LDAP provider that is not known to the metadata server's host.
Enable the metadata server itself to recognize the LDAP provider. See
“Direct LDAP Authentication” on page
You have accounts that are known at your web perimeter but are not known to the metadata server's host.
Enable the metadata server to trust users who have authenticated at the web
perimeter. See “Web Authentication” on page 149 .
Note: This is only a partial solution, because users of desktop applications still need accounts that can be validated by the metadata server or its host.
• In other cases, you must add accounts to your environment. Although it is technically possible to instead use SAS internal accounts for this purpose, those accounts are not intended for regular users.
• Anyone who directly connects to the OLAP server (without first connecting to the metadata server) needs an account with the OLAP server.
Note: Regardless of the location of your user accounts, you must also create
corresponding user information in the SAS metadata. Without such information, users have only the generic PUBLIC identity in the SAS realm. By default, this identity has no access to metadata and cannot log on to most applications. See
“Authentication to the Metadata Server” on page 114 .
How to Configure SAS Token Authentication
157
Coordinate the Workspace Server
Seamless access to the workspace server depends on coordination between that server and the metadata server. By default, this coordination is necessary because authentication to the workspace server is performed by the workspace server's host. The following table provides general recommendations:
Table 12.2 Coordinate the Workspace Server with the Metadata Server
Scenario
The servers run on
Windows or UNIX.
The servers run on z/OS.
The servers do not recognize the same accounts.
Recommendation
Use host authentication (either credential-based or Integrated Windows authentication). See
“Host Authentication” on page 137
or “Integrated Windows Authentication” on page 139 .
Use credential-based host authentication for both servers.
To minimize requirements for and exposure of host credentials, SAS provides several alternate
configurations. See “Mixed Providers” on page 117 .
Note: Similar coordination is not necessary for OLAP servers and stored process
servers, because they use SAS token authentication (instead of host authentication) for metadata-aware connections.
How to Configure SAS Token Authentication
Note: For metadata-aware connections to the stored process server, the pooled
workspace server, and the OLAP server, SAS token authentication is always used and no configuration is involved. For the standard workspace server, configuring
SAS token authentication is one of several solutions to a mixed provider situation.
The tasks in this section might have been automatically configured by the SAS
Deployment Wizard during deployment. If you chose to configure SAS token authentication at that time, no further configuration steps are required. If, however, you decided to set up SAS token authentication after installation, this section documents the steps to follow.
To configure the standard workspace server to use SAS token authentication, complete these steps:
1. Log on to SAS Management Console as someone who has user administration capabilities (for example, sasadm@saspw).
2. On the Plug-ins tab, expand Server Manager and the application server (for example, SASApp). Right-click the logical server (for example, SASApp -
Logical Workspace Server) and select Properties.
3. On the Options tab, select SAS token authentication. Click OK to save this setting.
4. Expand the logical server , select the server , right-click, and select
Properties.
158
Chapter 12 • Authentication Tasks
5. On the Options tab, from the Launch Credentials drop-down list, select a login and click OK. The most basic choice is the account for the SAS General Servers group
(the
sassrv
login).
6. To make the changes take effect, refresh the object spawner’s information about the server context. a. Connect to the object spawner server. Expand the object spawner , select the server , right-click, and select Connect. b. Refresh the object spawner server. Right-click the server, select Refresh
Spawner, and click Yes.
c. Disconnect from the object spawner server. Right-click the server and select
Disconnect.
7. To validate the configuration, select the logical workspace server , right-click, and select Validate.
Note: Metadata layer access from the server (for example, library pre-assignment,
PROC OLAP code, and metadata LIBNAME engine use) is still based on each requesting user's identity. However, all host access from the server is under the designated launch credential.
See Also
•
“SAS Token Authentication” on page 145
•
•
“Criteria for a Designated Launch Credential” on page 126
•
How to Configure Web Authentication
Overview of Configuring Web Authentication
Note: Before you configure web authentication, verify that this is an appropriate choice
in your environment. See “Web Authentication” on page 149 .
Configuring web authentication consists of the following high-level tasks:
1. Make configuration changes to SAS web applications (for example, add security information to the SAS Logon Manager application).
2. Make configuration changes to the SAS Web Application Server (for example, configure a realm to authenticate users against an LDAP or other source).
3. Rebuild and redeploy the SAS web applications. Update and restart the web application server.
4. Verify or adjust user information in the SAS metadata so that each user who participates in web authentication has an appropriate login in his or her metadata definition.
How to Configure Web Authentication
159
Step-By-Step Instructions for Web Authentication
For complete instructions for setting up web authentication, see
SAS Intelligence
Platform: Middle-Tier Administration Guide
.
Logins for Users Who Participate in Web Authentication
If you choose to configure web authentication, make sure that user metadata definitions include logins as explained in this topic.
Someone who uses only web applications should have a login in the web authentication domain. For example: web | joe | (no password)
Someone who uses both web and desktop applications might need two logins. One login contains the user's authenticated ID after logging on to a desktop application, and the other login contains the user's authenticated ID after logging on to a web application. For example:
DefaultAuth | WIN\joe | (no password) web | joe | (no password)
In the preceding example, two logins are needed because the format of the authenticated user ID differs in each context as follows:
• When Joe logs on to a desktop application (as joe), SAS obtains his user ID in downlevel format (WIN\joe), and that string is matched to the user ID in Joe's DefaultAuth login.
• When Joe logs on to a web application (as joe), SAS obtains his user ID in short format (joe), and that string is matched to the user ID in Joe's web login.
However, if the authenticated user ID is identical in both contexts, the web login is not needed. If SAS obtains both authenticated user IDs as joe, the web login is not needed.
In the following example, the metadata server is not authenticating against Windows accounts and the web login is not needed. When Joe logs on to a web application, the presence of his DefaultAuth login (which contains the correct user ID) is sufficient for the metadata server to successfully determine his metadata identity.
DefaultAuth | joe | (no password) web | joe | (no password)
Note: If your web environment uses Integrated Windows authentication, you must pay
careful attention to the format in which SAS obtains user IDs from the web realm. If you find that users of web applications have only the PUBLIC identity, it is likely that the user ID in each web login is not in the same format as the user ID that SAS obtains from the web realm.
Note: This is not a comprehensive discussion of logins; some users might have
additional logins for other purposes.
See Also
•
“How Logins Are Used” on page 121
•
“Authentication to the Metadata Server” on page 114
•
“Windows User ID Formats” on page 21
160
Chapter 12 • Authentication Tasks
How to Configure Direct LDAP Authentication
Before You Begin
Before you use these instructions, make sure that this is an appropriate choice in your
environment. See “Direct LDAP Authentication” on page 136 .
Active Directory
To make a metadata server on UNIX directly recognize Active Directory accounts, locate the sasv9_usermods.cfg file that is in your equivalent of
SAS/Config/Lev1/
SASMeta/MetadataServer
and add settings as follows:
/* Environment variables that describe your AD server */
-set AD_HOST myhost
/* System options that make AD the primary authentication provider */
-authpd ADIR:company.com -primpd company.com
After you restart the metadata server, the preceding settings cause these results:
Table 12.3 Example: User ID Formats If -authpd ADIR:company.com -primpd company.com
How a User Logs On
user
user@company.com
Where the Metadata Server
Sends the Credentials
To Active Directory
To Active Directory
How the User ID Must Be Stored in
Metadata
*
user@company.com
user@company.com
user@unrecognized-qualifier
WinDomain\user
user@saspw
user@host
To Active Directory
To Active Directory
To its internal provider
To its host
user@unrecognized-qualifier
WinDomain\user or user@WinDomain
No login for an internal account
user
*
If the ID is not stored in the correct format, the user can log on but has only the PUBLIC identity. Put the SAS copy of each user's ID in a login in that user's metadata definition. Assign these logins to DefaultAuth.
Note: To optimize credential reuse, do not move inbound logins (logins that provide
access to the metadata server) out of the DefaultAuth authentication domain.
Note: If you use external accounts for the SAS Administrator (sasadm) or the SAS
Trusted User (sastrust), certain configuration files that include those user IDs must conform to the format requirements in the third column of the preceding table. Only configuration files that contain the user ID for the purpose of matching an authenticated user ID must conform (for example, the adminUsers.txt and trustedUsers.txt files).
The following table provides reference information for environment variables that are used with the Active Directory provider.
How to Configure Direct LDAP Authentication
161
Table 12.4 Environment Variables for Active Directory
AD_HOST
AD_PORT
AD_STYLE
AD_TIMEOUT
The host name of the machine where Active Directory is running.
The port number for Active Directory. The default is 389.
Specifies one of the following user ID formats as the only format to be used in authentication.
"OLD" the down-level format (domain\user)
"UPN" the user principal name format (user@UPNsuffix)
"USERNAME" the short format (user)
Consider setting the AD_STYLE environment variable if user accounts are being locked due to exceeding the permitted number of unsuccessful log on attempts. If AD_STYLE is not set, as many as three authentication attempts are made (each format is tried sequentially, until an attempt succeeds or all formats fail). See
“Windows User ID Formats” on page 21 .
Specifies the number of seconds before time-out on the first attempt to connect. Secondary attempts use the default time-out (30 seconds).
AD_TLSMODE Enables Transport Layer Security (TLS) encrypted communication between the metadata server and the Active Directory server. Set this variable to 1 to activate (for example,
-set
AD_TLSMODE 1
).
*
*
Additional configuration is required. See
“How to Configure TLS between the Metadata Server and an LDAP Server” on page 207 .
Other LDAP Providers
To make a metadata server on UNIX or Windows directly recognize some other LDAP provider, locate the sasv9_usermods.cfg file that is in your equivalent of
SAS/Config/
Lev1/SASMeta/MetadataServer
and add settings as follows:
/* Environment variables that describe your LDAP server */
-set LDAP_HOST myhost
-set LDAP_BASE "ou=emp, o=us"
/* System options that make LDAP the primary authentication provider */
-authpd LDAP:company.com -primpd company.com
After you restart the metadata server, the preceding settings cause these results:
Table 12.5 Example: User ID Formats If -authpd LDAP:company.com -primpd company.com
How a User Logs On
user
user@company.com
user@unrecognized-qualifier
Where the Metadata Server
Sends the Credentials
To LDAP
How the User ID Must Be Stored in
Metadata
*
user@company.com
To LDAP
To LDAP
user@company.com
user@unrecognized-qualifier@company.com
162
Chapter 12 • Authentication Tasks
How a User Logs On
user@saspw
user@host
Where the Metadata Server
Sends the Credentials
To its internal provider
To its host
How the User ID Must Be Stored in
Metadata
*
No login for an internal account
user
*
If the ID is not stored in the correct format, the user can log on but has only the PUBLIC identity. Put the SAS copy of each user's ID in a login in that user's metadata definition. Assign these logins to DefaultAuth.
Note: To optimize credential reuse, do not move inbound logins (logins that provide
access to the metadata server) out of the DefaultAuth authentication domain.
The following table provides reference information for environment variables that are used with other LDAP providers.
Table 12.6 Environment Variables for Other LDAP Providers
LDAP_HOST The host name of the machine where LDAP is running.
LDAP_PORT
LDAP_BASE
The port number for LDAP. The default is 389.
The base DN to use. For example:
o=People, dc=orion, dc=com
.
LDAP_BASE_SUFFIXES Makes a blank-delimited set of domains (LDAP groups) available as suffixes for the
LDAP_BASE and LDAP_IDATTR environment variables. See “Multiple LDAP Groups” on page 163 .
LDAP_IDATTR (Optional) an alternative LDAP attribute that the SAS server can use to find your DN. The default is uid.
LDAP_PRIV_DN
The privileged DN that is allowed to search for users (for example,
cn=useradmin
).
*
LDAP_PRIV_PW The password for LDAP_PRIV_DN. You can use the PWENCODE procedure to provide an encoded password.
LDAP_TIMEOUT Specifies the number of seconds before time-out on the first attempt to connect. Secondary attempts use the default time-out (30 seconds).
LDAP_TLSMODE Enables Transport Layer Security (TLS) encrypted communication between the metadata server and the LDAP server. Set this variable to 1 to activate (for example,
-set
LDAP_TLSMODE 1
).
**
*
**
Set this variable only if users connect with a user ID instead of a DN, and the LDAP server does not allow anonymous binds.
Additional configuration is required. See
“How to Configure TLS between the Metadata Server and an LDAP Server” on page 207 .
Advanced Configurations
Multiple LDAP Servers
SAS can recognize multiple LDAP servers as authentication providers, regardless of whether there are trust relationships among those servers.
How to Configure Direct LDAP Authentication
163
For example, to cause the metadata server to search three Active Directory servers
(AD1, AD2, and AD3) when authenticating users, set the AUTHPROVIDERDOMAIN system option as follows:
-authpd (ADIR:AD1, ADIR:AD2, ADIR:AD3)
To enable SAS to locate all three LDAP servers, supply information about each server through associated environment variables. To indicate which LDAP server each variable describes, append the server's name to the variable name. For example, with the preceding AUTHPD syntax, you might set server-specific variables as follows: set AD_HOST_AD1=machine1.company1.com
set AD_HOST_AD2=machine1.company2.com
set AD_HOST_AD3=machine1.company3.com
T I P
Notice that the suffix of each server-specific variable corresponds to a value that is already set in the AUTHPD system option.
Note: From the perspective of a SAS server, loss of contact with a particular LDAP
server is not distinguishable from other possible causes of authentication failure
(such as invalid credentials). For this reason, configuring multiple LDAP servers does not provide failover support across those LDAP servers.
Multiple LDAP Groups
SAS can customize identity search behavior for different LDAP groups within an LDAP server. For each LDAP group, you can set a starting point for search (the LDAP base
DN) and designate an attribute on which to search (as an alternative to the default, which is to search by uid).
To enable SAS to differentiate among multiple LDAP groups, you first set an environment variable (LDAP_BASE_SUFFIXES) that registers the LDAP group names.
For example, if your LDAP user store has an LDAP group for each of two organizations
(EUROPE and AMERICAS), and you want SAS to use a different search strategy for each of those organizations, you set an environment variable as follows:
LDAP_BASE_SUFFIXES=EUROPE AMERICAS
You can then set additional environment variables to establish a base DN and an identity search attribute for each group. To indicate which LDAP group each variable should affect, you append the group name to the variable name, as depicted in the following partial example:
LDAP_BASE_EUROPE=OU=Europe...
LDAP_IDATTR_EUROPE=CN
LDAP_BASE_AMERICAS=OU=Americas ...
LDAP_IDATTR_AMERICAS=samaccountname
T I P
Notice that the suffix of each group-specific variable corresponds to a value that is already registered in the LDAP_BASE_SUFFIXES environment variable.
Like LDAP server-specific values, LDAP group-specific values are inherited if they are not explicitly set.
Inherited Values for Environment Variables
If a variable is not set at the server-specific level, SAS obtains a value for that variable from the corresponding shared variable. The preceding example assumes that all three servers are on the default port of their respective machines (389), so no AD_PORT variable is explicitly set. However, if all three of the servers were instead on the same
(non-default) port on their respective machines, you would add a shared AD_PORT= variable as follows:
164
Chapter 12 • Authentication Tasks
set AD_PORT=123 set AD_HOST_AD1=machine1.company1.com
set AD_HOST_AD2=machine1.company2.com
set AD_HOST_AD3=machine1.company3.com
If the third server was on another port, you would also set a server-specific variable as follows: set AD_PORT=123 set AD_HOST_AD1=machine1.company1.com
set AD_HOST_AD2=machine1.company2.com
set AD_HOST_AD3=machine1.company3.com
set AD_PORT_AD3=456
With the preceding settings, the AD1 and AD2 servers use port 123 on their respective machines. Because no server-specific port value is set for AD1 and AD2, those servers inherit the value from the shared AD_PORT variable. The AD3 server uses an explicitly set server-specific port value (456).
This pattern of inheriting shared values if there are no more specific settings applies to all of the AD_ and LDAP_ environment variables, except that any server that is listed as an LDAP value in the AUTHPD option must have a corresponding server-specific
LDAP_BASE_SUFFIXES variable.
Compound Suffixes for Environment Variables
You might need to append two suffixes to an environment variable if both of the following circumstances exist:
• You want SAS to directly authenticate users against multiple LDAP servers.
• You want SAS to use different search behavior for different LDAP groups within at least one of those LDAP servers.
If you need to specify two suffixes, append the LDAP server suffix first and the LDAP group suffix last (for example, LDAP_BASE_LDAPserver1_LDAPgroup1=value).
For example, suppose you specify the AUTHENTICATIONPROVIDERDOMAIN system option and associated environment variables as follows:
-authpd (LDAP:SERVER1, LDAP:SERVER2)
LDAP_HOST=hostname
LDAP_BASE=CN=Users,DC=us,DC=company,DC=com
LDAP_IDATTR=CN
LDAP_PRIV_DN=CN=useradmin
LDAP_PRIV_PW=pwd
LDAP_BASE_SUFFIXES_SERVER2=GROUPA GROUPB
LDAP_BASE_SERVER2_GROUPA=CN=Groupa,DC=us,DC=company,DC=com
LDAP_IDATTR_SERVER2_GROUPA=CN
LDAP_BASE_SERVER2_GROUPB=OU=GroupB,DC=us,DC=comnpany,DC=com
LDAP_IDATTR_SERVER2_GROUPB=CN
The results are as follows:
• SERVER1 uses all of the shared values.
• SERVER1 searches the base DN that is set by the shared LDAP_BASE variable.
• SERVER2 uses shared values for the LDAP_HOST, LDAP_PRIV_DN, and
LDAP_PRIV_PW variables.
How to Configure Integrated Windows Authentication
165
• SERVER2 searches within the base DN that is set in the
LDAP_BASE_CARYNT_GROUPA. If the identity is not found within that area,
SERVER2 also searches within the baseDN that is set in the
LDAP_BASE_CARYNT_GROUPB variable.
About AUTHPD and PRIMPD
Table 12.7 Reference: SAS System Options
AUTHPD
PRIMPD
Use this option to register and name your Active Directory provider or other LDAP provider. See
“AUTHPROVIDERDOMAIN System Option” in SAS System Options: Reference .
Use this option to designate your Active Directory server or other LDAP provider as the primary authentication provider for the metadata server. The metadata server directly uses its primary provider when the submitted user ID has no qualifier, the -primpd qualifier, or an unrecognized qualifier. Using this option enables users to log on using their usual user IDs (no special qualifier is required at logon time). See “PRIMARYPROVIDERDOMAIN= System Option” in SAS System Options: Reference .
About Setting Environment Variables
UNIX Specifics
On UNIX, an alternate location for specifying the environment variables is in the
MetadataServer.sh shell script. For example:
AD_HOST=myhost export AD_HOST
z/OS Specifics
On z/OS, a TKMVSENV file is used to make a list of pseudo environment variables available. A TKMVSENV PDS is created at installation. To define the environment variables, create a member in the PDS that specifies the necessary variables, and then reference this PDS member in the TKMVSENV DD statement in your started task.
For more information about setting environment variables, see the SAS system option
SET= in the documentation for your host:
•
SAS Companion for Windows
•
SAS Companion for UNIX Environments
•
SAS Companion for z/OS
How to Configure Integrated Windows
Authentication
Overview
Note: These instructions are for configuring Integrated Windows authentication (IWA)
from SAS desktop applications to the metadata server and the workspace server.
Before you configure IWA, verify that this is an appropriate choice in your environment. See
“Integrated Windows Authentication” on page 139
.
166
Chapter 12 • Authentication Tasks
Configuration of IWA for desktop applications can involve three distinct locations:
• Client participation in IWA is determined by a setting in each client-side connection profile. If IWA is not selected by a client, IWA is not used for that client.
• Server participation in IWA is affected by invocation commands. For example, to use
IWA, the metadata server and object spawner start commands must include the option
-sspi
.
Note: This is the default on Windows and UNIX
• For metadata-aware connections to a workspace server, participation in IWA is also affected by settings in that server's metadata definition.
Instructions
1. If the metadata server or workspace server runs on UNIX, complete the UNIX prerequisite tasks. Before you can use IWA for a SAS server that runs on a UNIX host, you must prepare and configure the UNIX environment. For example:
• You must acquire, install, and configure the required software. For the first maintenance release for SAS 9.4, the only supported implementation of IWA on
UNIX requires Quest Authentication Services (QAS) 4.0.1.23 or later. For the second maintenance release for SAS 9.4 on Linux systems and the fourth maintenance release for SAS 9.4 on Solaris systems, any shared library
(including QAS) that implements the GSSAPI with Kerberos 5 extensions can be used. If QAS is present on the host, SAS attempts to load its libraries before attempting to load others to retain legacy behavior.
Note: In a single deployment, do not use Quest with other shared libraries.
• You must create a service account and corresponding keytab file. Participating
SAS processes on UNIX must be able to read the keytab file. A keytab file is functionally equivalent to a user's password and should be secured in the same way. The keytab file should be owned by the user and group that is accessing it, and it should have the least permissive access mask (preferably 0600).
• You must set certain environment variables.
These prerequisite tasks should be performed during the installation and initial configuration phases of your deployment. Some of the implementation details differ by UNIX host. For these reasons, detailed instructions for preparing your UNIX host to support IWA are in a separate document. See the chapter "Configuring Integrated
Windows Authentication" in Configuration Guide for SAS Foundation for UNIX
Environments at
http://support.sas.com/documentation/installcenter .
2. If the metadata server is clustered and runs on Windows, or if your SAS servers are configured using DNS aliases, manually register SPNs. See
“Manual Registration” on page 169
.
3. Verify an IWA connection to the metadata server.
a. In a client-side connection profile, select the check box that enables Integrated
Windows authentication and then attempt to connect.
T I P
For example, in the Connection Profile dialog box in SAS Management
Console, click Edit. On the Connection Information page of the Edit
Connection Profile wizard, select the Use Integrated Windows
authentication (single sign-on) check box.
b. If the connection fails, verify the following:
How to Configure Integrated Windows Authentication
167
• The metadata server's start-up command includes
-sspi
.
• The advanced IWA settings in your client-side connection profile are
Negotiate for the security package, blank (no value) for the service principal name, and Kerberos,NTLM for the security package list.
• Your metadata user definition includes a login that contains your user ID in
the correct format. See “User Logins” on page 168
.
• You are using a Windows desktop client (the SAS implementation of IWA is not for web applications or UNIX clients).
• If the metadata server runs on UNIX, the prerequisite steps have been successfully completed. (See step 1 above.) c. After the connection succeeds, examine the metadata server log to confirm that an IWA connection was used. If a credential-based connection occurred, make sure that your password is not cached in the client application.
T I P
In SAS Management Console, you can clear the credentials cache by selecting File
ð
Clear Credentials Cache from the main menu.
4. Configure the workspace server’s metadata definition for IWA, and verify an IWA connection to that server.
T I P
If you use IWA for the metadata server, there are no cached credentials from an initial logon. For this reason, it is a good idea to configure IWA for the standard workspace server also, if possible (IWA is not supported on z/OS).
a. On the Plug-ins tab in SAS Management Console, expand Server Manager and the application server (for example, SASApp). Right-click the logical server
(for example, SASApp - Logical Workspace Server) and select Properties.
b. On the Options tab, select the Host radio button (IWA is a form of host authentication).
• Select the Negotiate security package.
• Leave the Security package list as Kerberos,NTLM.
• Leave the Service principal name blank. In a standard configuration, clients expect (and know how to compose) the default SPN. Entering a value here
(or on the client side) overrides this default process.
c. Log on to SAS Management Console using IWA. Right-click the Logical
Workspace Server and select Validate.
d. If the connection fails, verify the following:
• The object spawner's start-up command includes
-sspi
.
• Your metadata user definition includes a login that contains your user ID in
the correct format. See “User Logins” on page 168
.
• If the workspace server runs on UNIX, the prerequisite steps have been successfully completed. (See step 1 above.) e. After the connection succeeds, examine the object spawner log to verify that the connection to the workspace server was made using IWA. If the spawner log indicates that credential-based authentication occurred (instead of IWA), the user's context includes credentials for the workspace server's host. Make sure that the user does not have a cached or stored password for the workspace server’s authentication domain.
Note: Even if IWA is configured, any available cached or stored credentials are
preferentially used.
168
Chapter 12 • Authentication Tasks
5. If the workspace server needs to access Kerberized network resources (such as network file systems or IWA connections to databases):
• Edit the Security package list so that only Kerberos is specified.
• In Active Directory, make the object spawner account trusted for delegation to all services. See
“Windows Privileges” on page 19
.
6. Inform users that they can select the IWA option when they log on to desktop applications such as SAS Information Map Studio, SAS Data Integration Studio,
SAS OLAP Cube Studio, SAS Management Console, and SAS Enterprise Guide. In general, users should not make changes to the advanced IWA settings in their clientside connection profiles.
See Also
• “Checking the Status of Servers” in SAS Intelligence Platform: System
Administration Guide
• “Default Locations for Server Logs” in SAS Intelligence Platform: System
Administration Guide
• “Spawner Invocation Options” in SAS Intelligence Platform: Application Server
Administration Guide
User Logins
If you choose to configure IWA, make sure that user metadata definitions include logins with properly formatted user IDs. The format of the stored user IDs must match the format in which authenticated user IDs are returned to the target server. Failure to meet this requirement causes the user to have only the generic PUBLIC identity (which, by default, cannot even log on to most applications).
In the standard configuration, the appropriate format varies as follows:
• If the target server is on Windows, the authenticated user ID is returned in qualified format, so the stored user ID should be qualified (for example, WIN\joe or
• If the target server is on UNIX, the authenticated user ID is returned in short format
(it is not qualified), so the stored user ID should not be qualified (for example, joe or
fred.smith).
If you need to align formats, use the SASUSEKERBNAME environment variable. For example, you might use this environment variable in either of the following circumstances:
• The metadata server is on Windows, the workspace server is on UNIX, both are using IWA, and you do not want to store two logins for each user.
• You need to distinguish between two different users, in two different Kerberos realms, who happen to have the same sAMAccountName name (for example, [email protected] and [email protected]).
See
“Windows User ID Formats” on page 21
.
How to Configure Integrated Windows Authentication
169
Registering SPNs
Automatic Registration
In a standard configuration on Windows, you do not manually register service principal names (SPNs). Instead, each participating SAS server uses the local system account under which it runs to self-register two SPNs against the computer object of the machine where it runs. These Active Directory registrations are in the format SAS/fully-qualified-
machine-name and SAS/netbios-machine-name (for example,
SAS/machineA.company.com and SAS/machineA).
In a standard configuration on Windows, clients can construct a standard SPN (because they know the format and the target machine), so users do not have to supply an SPN when they initiate an IWA connection.
Manual Registration
In circumstances such as the following, manual registration of SPNs is necessary in order to support Kerberos connections:
• Your metadata server is clustered and runs on Windows. See
.
• Your SAS servers use DNS aliases, so the SPNs must reference an alias instead of a machine name. See
“IWA to Servers That Use DNS Aliases” on page 174 .
• You are using IWA for servers on UNIX systems.
CAUTION:
SPNs must be unique among the objects that are in the same Active Directory forest.
A given SPN must not be assigned to more than one object. If a duplicate
SPN is found in response to a TGS request, the Key Distribution Center sends an error to the client that the principal was not found.
To manually register an SPN, use the Microsoft tool setspn . For example, the following command registers
customValue
as the SPN for all servers that run as services under the local system account on a machine that is named myServer: setspn -A customValue myServer
Note: You must be a Windows domain administrator in order to use the
setspn
command.
Note: You must also make any necessary adjustments to client-side connection profiles
(and the logical workspace server definition, if applicable). For example, if you supplied a custom value for the SPN, specify the new customValue in the Service
principal name fields.
UNIX Specifics
On UNIX, the SPNs that are used must be listed in the keytab file. In addition to running
setspn
to set a custom SPN, and making sure that client connection profiles use that custom SPN, you must generate a new keytab file that includes the new SPN. See the chapter "Configuring Integrated Windows Authentication" in
Configuration Guide for SAS Foundation for UNIX Environments.
Manually Registering Object Spawner SPNs
Note: The following manual registration is required only on Windows systems. It does
not have to be completed on UNIX systems.
170
Chapter 12 • Authentication Tasks
By default, the SAS Object Spawner service runs as a local system. The best practice is to set this service to run under a service level account. Although the account must be an administrator, it allows system administrators to allow delegation to a single account rather than the whole computer if, delegation must be set to allow access to third-party resources (such as SQL and network shares).
When using a service level account to run the object spawner service in a SAS Grid environment, you need to configure the default SPNs: setspn –A SAS/computerNetbios –u domain\ObjectSpawnerServiceAccount setspn –A SAS/computerFullname –u domain\ObjectSpawnerServiceAccount
In non-grid environments, you can configure custom SPNs, such as the following: setspn –A SASWS/computerNetbios –u domain\ObjectSpawnerServiceAccount setspn –A SASWS/computerShortname –u domain\ObjectSpawnerServiceAccount
Note: The preceding SPN statements differ from the one that is built if you use the local
system.
Manually Registering SQL SPNs
When registering a SQL Server database to use IWA, additional configuration steps are needed. The following scenarios provide information about possible configurations.
In the first scenario, the SQL Server service runs under a service account and SQL
Server service is not on the same machine as the object spawner service. For this scenario, follow these steps:
1. To register custom SPNs for the service, run the following commands: setspn –A MSSQLSVC/computerNetboisName: sqlPort –u domain\sqlServiceAccount setspn –A MSSQLSVC/fullyQualifiedDomainName: port –u domain\sqlServiceAccount setspn –A MSSQLSVC/machineName: sqlInstance –u domain\sqlServiceAccount
Create the SPNs for each DNS alias, if any are being used.
2. Run the
setspn –x
command to verify that no duplicate SPNs exist.
3. Verify that the SQL Server is accepting Kerberos connections. Run the following command from the SQL Management Studio on the computer where the object spawner service runs:
SELECT auth_scheme FROM sys.dm_exec_connections WHERE session_id = @@spid;
If the query does not return KERBEROS, the SPNs might not be created correctly, or there might be other issues, such as port 88 (the Kerberos ticketing service) is blocked.
4. Create an ODBC connection on the same machine where the spawner resides, by completing the following steps: a. From the Windows Start menu, select Control Panel
ð
Administrative Tools
ð
Data Sources (ODBC). The ODBC Data Source Administrator dialog box is displayed.
b. Select the System DSN tab and click Add.
c. Use the Create New Data Source wizard to configure the connection.
In the second scenario, the SQL Server service runs under a service account and is on the same machine as the object spawner service. In this scenario, since both services reside
How to Configure Integrated Windows Authentication
171
on the same machine, no delegation is required for the object spawner to access the SQL
Service.
Note: You need to configure delegation if the object spawner needs to delegate
credentials across the network, for example, to access a network share.
In the third scenario, the SQL Server service runs under a local system account and is on the same machine as the object spawner service. This configuration is not a best practice and is therefore not supported by Microsoft. If a custom SPN for the SQL Server service account is required or preferred, the Microsoft SQL Server 2012 Native Client needs to be installed and configured.
Forcing Kerberos
If you choose to use the SAS implementation of Integrated Windows authentication
(IWA), and you need to ensure that the Kerberos protocol is always used, follow these instructions. These instructions assume that you have already fully configured IWA.
Note: You cannot use Windows local accounts with this configuration, because those
accounts cannot use Kerberos.
1. Specify
-secpackagelist "Kerberos"
in your equivalent of the following locations:
•
SAS\Config\Lev1\SASMeta\MetadataServer\sasv9_usermods.cfg
(for the metadata server)
•
SAS\Config\Lev1\SASApp\OLAPServer\sasv9_usermods.cfg
(if you need to support direct IWA connections to an OLAP server on Windows)
•
SAS\Config\Lev1\ObjectSpawner\ObjectSpawner.bat
(if the object spawner is on Windows). If the spawner runs as a service, complete these steps:
1. From the Windows Start menu, select Programs
ð
SAS
ð
SAS
Configuration
ð
<Level>
ð
Object Spawner – Stop.
2. From the object spawner's configuration directory, enter the following:
ObjectSpawner.bat -remove
3. Add the
-secpackagelist "Kerberos"
setting to the
Set
CMD_OPTIONS=
line of ObjectSpawner.bat. Also, make sure that the
-sspi
setting is present.
4. To reinstall the spawner service, enter the following:
ObjectSpawner.bat -install
2. Restart the metadata server.
3. Make sure that the workspace server’s metadata definition includes only Kerberos in the Security package list field. This setting is located in SAS Management
Console, on the Plug-ins tab, under Server Manager. The setting is on the Options tab of the logical workspace server definition.
Note: On Windows systems, if the workspace server accesses network resources
(such as network filesystems or IWA connections to databases), you must also mark the account under which the spawner runs as
trusted for delegation to all services
. See
“Windows Privileges” on page 19
.
Note: On UNIX systems, the account that the SAS SPN is registered against must be
trusted for delegation in Active Directory.
4. Restart the object spawner.
172
Chapter 12 • Authentication Tasks
T I P
In general, it is not necessary to also change the default IWA setting in clientside connection profiles. If a server accepts only Kerberos, then clients with the default setting of Negotiate (and both Kerberos and NTLM in the security package list) use Kerberos. However, in some circumstances, the client's Windows system chooses to initiate communication using NTLM and is unable to comply with the server requirement by switching to Kerberos. For example, if the client and server are on the same machine, the client chooses NTLM. In these circumstances, you must adjust the client-side settings to specify only Kerberos.
UNIX Specifics
It is not necessary to force use of Kerberos on UNIX, because IWA on UNIX supports only the Kerberos protocol.
Supported Protocols
IWA requires agreement between client and server about which security protocol to use when exchanging authentication packets. The following table provides details:
Table 12.8 Integrated Windows Authentication Settings
Server
Setting Associated Requirements
Negotiate security package
• The client must select the Negotiate security package.
• The server must have a security package list. By default, servers have a security package list that offers the Kerberos and NTLM protocols.
• At least one of the protocols in the server's security package list must be offered by the client.
• At least one of the protocols that are offered by both parties must actually be supported by both parties.
The NTLM protocol is not supported on UNIX.
• The Kerberos protocol can be used only if the client knows the server's SPN, as explained in the following row.
Kerberos security package
• The client must select the Kerberos security package.
• Both client and server must actually support the Kerberos protocol.
• The client must know the server's service principal name (SPN). In a standard configuration, this is transparent. Clients expect (and know how to construct) a default SPN in the format
SAS/
machine (for example, SAS/machineA.na.company.com), so you do not have to explicitly provide the SPN.
Note: For a server on UNIX, the machine value must be specified as a fully qualified domain name
(FQDN). For a server on Windows, you can instead specify the machine’s NetBIOS name, but the
FQDN is preferred (because a NetBIOS name is not guaranteed to be unique).
NTLM security package
• The client must select the NTLM security package.
• Both client and server must actually support the NTLM protocol. The NTLM protocol is not supported on UNIX.
IWA to an OLAP Server
In the standard configuration, the OLAP server supports IWA for direct connections (for example, from an open OLAP client that uses an OLE DB Provider for OLAP). This support is similar to the metadata server’s support of IWA for direct connections from
How to Configure Integrated Windows Authentication
173
SAS desktop clients. In both cases, client-side connection information must request that
IWA is used.
Note: In the platform, most connections to the OLAP server are metadata-aware, not
direct (the client first connects to the metadata server and then connects to the OLAP server, rather than initially connecting to the OLAP server). Metadata-aware connections to the OLAP server use SAS token authentication (they do not use
IWA).
IWA to a Clustered Metadata Server
Introduction
A clustered metadata server on Windows runs under a domain account. No local system account is available to perform the usual self-registration of SPNs. Instead, you must manually register SPNs for all of the host machines in the cluster.
Example
For a metadata server that runs as [email protected] on three dedicated machines
(machineA, machineB, and machineC), you might register the following SPNs:
• SAS/machineA.company.com
• SAS/machineA
• SAS/machineB.company.com
• SAS/machineB
• SAS/machineC.company.com
• SAS/machineC
In this example, the SPNs are registered against the Active Directory principal that
corresponds to the [email protected] account. See “Manual Registration” on page
Shared Hardware
If other SAS servers that run as services share any of the clustered metadata server’s hardware, Active Directory conflicts can occur, because one SPN cannot be associated with two Active Directory identities.
You can work around such conflicts by using either of the following methods:
• Reconfigure the other servers to run under the metadata server’s domain account
(instead of running under the local system account).
Note: If you make this adjustment to the object spawner, the domain account must
be a Windows administrator on the spawner's host and must have the Windows user rights Adjust memory quotas for a process and Replace a process level
token. These user rights assignments are part of the local security policy for the
Windows computer that hosts the spawner.
• Adjust the configuration to reference the domain account’s user principal name
(UPN). For example, if the metadata server runs as [email protected], and the
UPN that is registered in Active Directory for that account is [email protected], then users should enter the value [email protected] in the Service principal name (SPN) field of the IWA Advanced Settings dialog box in their client-side connection profiles.
174
Chapter 12 • Authentication Tasks
IWA to Servers That Use DNS Aliases
To support Kerberos connections to a server that uses a DNS alias, manually register
SPNs that reference the alias (instead of referencing the machine name).
For example, for a metadata server on Windows that runs under the local system account and uses the alias DNSalias, you might register the following custom SPNs.
• SAS/DNSalias
• SAS/DNSalias.company.com
In this example, the SPNs are registered against the Active Directory computer object
that corresponds to the machine on which the metadata server runs. See “Manual
About IWA from SAS Web Applications
The SAS implementation of IWA is for desktop applications only. Web applications can use IWA if they are configured for web authentication and the web application server supports IWA. See SAS Intelligence Platform: Middle-Tier Administration Guide.
How to Store Passwords for the Workspace
Server
Note: This is one of several solutions to a mixed provider situation (and it is not a
preferred approach). See
“Mixed Providers” on page 117 .
1. Log on to SAS Management Console as someone who has user administration capabilities (for example, sasadm@saspw).
2. On the Plug-ins tab, expand Server Manager, the application server (for example, SASApp) and the logical server (for example, SASApp - Logical
Workspace Server).
(Optional) Right-click the logical server, select Properties, and select the Options tab. Make sure that the Authentication service is set to use Host with Username/
Password. Click OK to close the dialog box and return to Server Manager.
3. Below the logical server , select the server (for example, SASApp -
Workspace Server).
4. In the display panel, right-click the server's connection object and select
Properties.
See Also
How to Store Passwords for a Third-Party Server
175
5. On the Options tab, notice the value in the Authentication domain drop-down list.
If the value is something other than DefaultAuth, proceed to step 6. Otherwise, complete these steps: a. Next to the Authentication domain drop-down list, click New. b. In the New Authentication Domain dialog box, enter a name such as
UNIXAuth
or
WorkspaceAuth
(you can use any name that is meaningful to you). Click
OK.
c. On the Options tab, make sure the new authentication domain is selected in the
Authentication domain drop-down list. Click OK.
6. Create a SAS copy of credentials that are known to the workspace server's host operating system. In most cases, you store shared credentials in a group's metadata definition. You can also store a unique set of individual credentials in each user's metadata definition. Each login must be assigned to the workspace server's authentication domain. Each login must include both a user ID and a password.
Here is an example for a workspace server on UNIX:
UNIXAuth | myID | mypassword
Here is an example for a workspace server on Windows:
WINAuth | Win\myID | myWINpassword
If you store credentials for a workspace server that runs on Windows, give users the
Windows privilege Log on as a batch job.
Note: If you do not store the passwords, users are prompted for such credentials
when they make a request that requires access to the workspace server. Only desktop applications and SAS Web Report Studio provide such secondary logon prompts.
Note: Do not instead leave the workspace server in DefaultAuth and move inbound
logins to some other authentication domain. Failure to follow this recommendation does not affect the initial logon process, but it does interfere with access to the workspace server. By default, all clients insert the credentials that a user submits at the logon prompt into that user's context as a DefaultAuth entry. This cached
DefaultAuth entry has priority over any DefaultAuth credentials that are retrieved from logins in the metadata.
•
“Credential Management” on page 134
•
“How Logins Are Used” on page 121
•
“Windows Privileges” on page 19
How to Store Passwords for a Third-Party Server
Note: Use these instructions to provide seamless access to a third-party server that uses
a proprietary authentication provider (for example, Oracle). These instructions associate the database logins with a user or group, not directly with a database library.
176
Chapter 12 • Authentication Tasks
1. Verify that the third-party server is registered in the metadata and is in its own authentication domain. a. Select the third-party server's definition under Server Manager on the Plug-ins tab in SAS Management Console.
b. In the display panel, right-click the server's connection object and select
Properties. The server's authentication domain assignment is on the Options tab.
2. In the server's authentication provider, identify or create accounts. Use any of the following approaches (here, Oracle is used as an example):
• Create an individual Oracle account for each user. This provides the greatest accountability, but can also necessitate storing many Oracle user IDs and passwords in the metadata.
• Create one Oracle account that all users share. This greatly reduces the need to store Oracle user IDs and passwords, but also results in a loss of individual accountability.
• Create a few Oracle accounts, each of which is shared by several users. This middle-of-the-road approach enables you to make some access distinctions in
Oracle and store only a few Oracle user IDs and passwords in the metadata.
3. In the metadata, store the user IDs and passwords for each account as follows (here,
Oracle is used as an example):
• If you created individual accounts on the Oracle server, add an Oracle login to each user definition.
• If you created one shared account on the Oracle server, identify or create a group that contains the users who access the Oracle server. Give that group a login that includes the user ID and password for the Oracle shared account.
Note: If you want to provide access for all registered users, give the login to the
SASUSERS group.
Note: If you want to provide access for all users (including users who do not
have an individual SAS identity), give the login to the PUBLIC group.
• If you created several shared accounts on the Oracle server, identify or create a user group in the metadata for each shared account. Give each group a login for the Oracle server, and assign each user who connects to Oracle to one of the groups.
Assign these logins to the third-party server's authentication domain. Store both an
ID and a password in each login.
Here is an example for an Oracle server:
OracleAuth | myORAid | myORApassword
Note: If you do not store the passwords, users of desktop applications are prompted
for such credentials when they make a request that requires access to the server.
SAS Web Report Studio has an interactive password management feature. Other web applications do not support interactive logons to secondary servers.
See Also
•
“Credential Management” on page 134
•
“How Logins Are Used” on page 121
How to Change Internal Account Policies
177
How to Change Internal Account Policies
Server-Level Policies
Initial Settings
Here are the initial server-level policies for internal accounts:
• Accounts do not expire and are not suspended due to inactivity.
• Passwords must be at least six characters, do not have to include numbers or mixed case, and do not expire.
• The five most recent passwords cannot be reused.
• After three failed attempts to log on, an account is locked for one hour. An administrator can unlock the account by accessing the Accounts tab in the user's definition in SAS Management Console.
• A forced password change occurs on first use and after a password is reset. This policy applies only to accounts with passwords that periodically expire. By initial policy, passwords do not expire, so forced password changes do not occur.
Current Settings
You can examine the current policies in SAS Management Console. On the Plug-ins tab, expand the Metadata Manager, right-click the Active Server node, and select
Properties. The Options tab displays any server-level internal account policies that have been explicitly set.
Note: You can also examine explicitly set policies by viewing the omaconfig.xml file.
Syntax for the InternalAuthenticationPolicy Element
To change the server-level policies, edit the
InternalAuthenticationPolicy
element in the metadata server's omaconfig.xml file, and then restart that server.
Here is the syntax for each policy option:
Note: The following option names are case-sensitive.
Note: A value of
T
has aliases (
1
or
Y
). A value of
F
has aliases (
0
or
N
).
ChangeDelayInMinutes=
"number" specifies the number of minutes that must elapse between password changes. Applies only when you are resetting your own password.
DigitRequired="T | F"
specifies whether passwords must include at least one digit. To enforce this requirement, specify
T
.
ExpirationDays=
"number" specifies the number of days after password is set that the password expires. A value of 0 prevents passwords from expiring.
ExpirePasswordOnReset="T | F"
specifies whether a forced password change occurs on first use and after an administrative password reset. To disable this requirement, specify
F
.
178
Chapter 12 • Authentication Tasks
HashPasswords="SHA256 | MD5"
specifies how the internal account password is stored in the metadata.
SHA256 the SHA-256 hash function is used. SHA (secure hash algorithm) is
FIPS (Federal Information Processing Standard) compliant.
SHA-256 hashing is provided by SAS/SECURE. This is the default.
MD5 MD5 hashing is used. MD5 (message digest algorithm 5) is appropriate for preventing accidental exposure of information.
MinLength=
"number-of-characters" specifies the minimum length for passwords.
MixedCase="T | F"
specifies whether passwords must include at least one uppercase letter and at least one lowercase letter. To enforce this requirement, specify
T
.
NumPriorPasswords=
"number" specifies the number of passwords that are maintained in each account's password history. A user cannot reuse a password that is in the user's account history.
InactiveDaysToSuspension=
"number" specifies the number of days after which an unused account is suspended. A value of
0 prevents suspensions due to inactivity.
Note: Beginning in the fourth maintenance release for SAS 9.4, this option is no
longer supported.
LockoutDurationInMinutes=
"number" specifies the number of minutes for which an account is locked following excessive login failures.
NumFailuresForLockout=
"number" specifies the number of consecutive unsuccessful logon attempts that cause an account to be locked. We recommend that you do not specify 0, because doing so can make your system vulnerable to password guessing attacks.
Example of the InternalAuthenticationPolicy Element
<OMAconfig>
...
<InternalAuthenticationPolicy ChangeDelayInMinutes="0" DigitRequired="F"
ExpirationDays="0" MinLength="6" MixedCase="F" NumPriorPasswords="5"
InactiveDaysToSuspension="0" LockoutDurationInMinutes="60"
NumFailuresForLockout="3"/>
...
</OMAconfig>
Per-Account Policies
To override server-level policies on a per-account basis:
1. Log on to SAS Management Console as someone who has user administration capabilities.
2. On the Plug-ins tab, select User Manager.
3. In the display pane, clear the Show Groups and Show Roles check boxes. Rightclick the user definition of the user whose SAS internal account policies you want to change. Select Properties.
4. At the bottom of the user's Accounts tab, click Update.
About the Workspace Server's Options Tab
179
5. Make changes in the Custom Settings box. Not all server-level settings can be modified on a per-account basis.
Note: There are two distinct expiration settings. Do not confuse the account
expiration date with the password expiration period.
Note: To minimize administrative maintenance effort for any predefined or service
identities that have internal accounts, do not add expiration dates to these accounts or expiration periods to these passwords.
The following table maps server-level policies to corresponding account-level policies. Not all policies can be set at both levels.
Table 12.9 Internal Account Policy Mapping
Server-Level Policy
ExpirationDays
LockoutDurationinMinutes
NumFailuresForLockout
NumPriorPasswords
Related Account Level Setting
Set a custom password expiration period.
Exempt from account lockout policy.
Exempt from account lockout policy.
Exempt from password reuse policy.
T I P
If you want to force a particular user to change his or her internal password after you create (or reset) the user's internal account, but you do not otherwise want the password to expire, use the maximum custom password expiration period (32,767 days).
See Also
“SAS Internal Authentication” on page 143
About the Workspace Server's Options Tab
The following tips help you interpret the Options tab for a logical workspace server in SAS Management Console.
• Changes that you make on this tab take effect after you refresh the object spawner.
• Because the displayed settings are in the metadata, they can affect only metadataaware connections.
• The Use Server Access Security check box should be selected. This enables standard access control enforcement for the ReadMetadata permission on the server definitions.
Note: An exception is if the workspace server is configured for client-side pooling.
In that case, the check box is disabled and cleared. Server access security is not supported with client-side pooling.
• The Host with Username/Password setting specifies that credential-based host authentication is always used. The other security packages cause IWA to be used for clients that select the IWA option (unless cached or stored credentials are available).
180
Chapter 12 • Authentication Tasks
• The Host with Username/Password setting does not eliminate prompting for credentials. With this setting, all desktop applications prompt users to interactively supply credentials in any circumstance where credentials are needed and are not otherwise available.
• If you select SAS token authentication, you must also select a server launch credential (on the Options tab of the server definition). Configuring a workspace server to use SAS token authentication can be useful in a multi-host environment.
See “Mixed Providers” on page 117 .
• The Host with Username/Password setting can cause SAS Enterprise Guide and the
SAS Add-In for Microsoft Office to silently store user passwords in metadata. The storage occurs only if the workspace server is in its own authentication domain (for example, ServerAuth) and users interactively provide credentials when they access that server. If both of these circumstances apply to your deployment, consider selecting Prompt instead of Host with Username/Password.
• The Prompt setting is similar to Host with Username/Password, but it forces SAS
Enterprise Guide and the SAS Add-In for Microsoft Office to prompt users.
See Also
•
“Integrated Windows Authentication” on page 139
•
“SAS Token Authentication” on page 145
181
Part 4
Encryption
Chapter 13
Encryption Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Chapter 14
Encryption Tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
182
183
Chapter 13
Encryption Model
Managing Certificates in the SAS Environment . . . . . . . . . . . . . . . . . . . . . . . . . . 190
Certificate Formats Supported by SAS Middle-Tier Components . . . . . . . . . . . . . 192
How Many Private Keys and Signed Certificates Do I Need? . . . . . . . . . . . . . . . . 194
Introduction to the Encryption Model
Two Classes of Encryption Strength
Two classes of encryption strength are available:
• For compatibility with legacy systems, SASProprietary encoding is supported. This method is available in all deployments, is appropriate for preventing accidental exposure of information, and has minimal impact on performance.
• For a higher level of security, it is recommended to use industry-standard encryption and hashing algorithms. These methods provide stronger protection and are available in all deployments, except where prohibited by import restrictions.
Note: Industry-standard algorithms are provided by SAS/SECURE. For details
about supported algorithms and availability, see Encryption in SAS.
SAS recommends that you use the strongest security standards available for your environment.
184
Chapter 13 • Encryption Model
Two Contexts for Encryption Coverage
SAS provides encryption in two contexts:
• In encryption for data at rest, the emphasis is on protection of passwords in configuration files and in the metadata repository. You can also choose to encrypt
SAS data sets.
• In encryption for data in motion, the emphasis is on protection of passwords in transit. You can also choose to protect all traffic in transit among SAS Integrated
Object Model (IOM) servers and SAS desktop clients.
Note: To ensure that only FIPS-validated encryption algorithms are used, set the
ENCRYPTFIPS system option. See “ENCRYPTFIPS System Option” in
Encryption in SAS
.
See Also
•
“Encryption for Data at Rest” on page 184
•
“Encryption for Data in Motion” on page 185
Encryption for Data at Rest
In general, encryption for data at rest is set up during installation and no further adjustments are required. The following table lists supported algorithms by storage context.
Table 13.1 Encryption for Data at Rest
Storage Context
Login password on disk in the metadata
Internal account password on disk in the metadata
Password on disk in a configuration file
Supported Algorithms
SASProprietary encoding or AES
MD5 or SHA-256 hashing
SASProprietary encoding or AES
Note: If your configuration specifies an algorithm that is not available, the
SASProprietary algorithm and MD5 hashing are used instead. See
Encryption Strength” on page 183 .
Note: Most other metadata is not stored in an encrypted format. Configuration files and
metadata repository data sets should also be host protected. See “First-Priority Setup
Tasks” in SAS Intelligence Platform: Installation and Configuration Guide .
T I P
To make any post-installation changes, see
“How to Maximize Encryption of
Stored Passwords” on page 198 .
T I P
To encrypt SAS data sets on disk, see “ENCRYPT= Data Set Option” in SAS
Data Set Options: Reference
.
Encryption for Data in Motion
185
T I P
To require encryption of the tables in a metadata-bound library and store the encryption key in metadata, see the
SAS Guide to Metadata-Bound Libraries
.
Encryption for Data in Motion
In general, encryption for data in motion settings are defined during installation and no further adjustments are required. The following table lists supported algorithms by content and network traffic context. The table also references instructions for making post-installation changes.
Table 13.2 Encryption for Data in Motion
Content and Context Supported Algorithms Reconfiguration
Password in transit in a logon attempt
(from a SAS IOM client, data provider, or Base session to a SAS
IOM server)
SASProprietary encoding, RC2,
RC4, DES, TripleDES, or AES
See
“How to Change Encoding for Data in
Motion for IOM Servers” on page 202
.
General traffic among SAS IOM servers and SAS IOM clients
None, SASProprietary encoding, RC2, RC4, DES,
TripleDES, or AES
*
Password in transit from a client to the metadata server (when the client stores a login in metadata)
SASProprietary encoding
Password in transit from the metadata server to a client (when the client retrieves a stored login)
SASProprietary encoding or
AES
**
Not configurable.
See
“How to Change Encoding for Data in
Motion for Outbound Passwords” on page
.
Traffic from the metadata server to an
LDAP server
None or TLS
Traffic between the SAS Web Server and the SAS Web Application Server
None or TLS
See
“How to Configure TLS between the
Metadata Server and an LDAP Server” on page 207 .
See
SAS Intelligence Platform: Middle-Tier
Administration Guide
.
Traffic between a web browser and the
SAS Web Server
None or TLS
SAS Deployment Agent traffic None or TLS
SAS Environment Manager agent-toserver traffic
None or TLS
SAS Environment Manager server-toagent traffic
TLS
See
SAS Intelligence Platform: Installation and Configuration Guide
.
See SAS Environment Manager: User’s
Guide.
*
**
The algorithm that you choose is used if the NETENCRALG option is set to EVERYTHING.
These algorithms encode the password in the transmitted XML. In addition, the response traffic is encrypted using the algorithm that is specified in the NETENCRALG option.
186
Chapter 13 • Encryption Model
Note: If your configuration specifies an algorithm that is not available, the
SASProprietary algorithm is used instead. See “Two Classes of Encryption Strength” on page 183
.
What Are Certificates?
TLS and its predecessor, Secure Sockets Layer (SSL), are cryptographic protocols that are designed to provide communication security over the Internet. TLS and SSL are protocols that provide network data privacy, data integrity, and authentication.
Note: All discussion of TLS is applicable to the predecessor protocol, SSL.
TLS uses X.509 digital certificates and asymmetric cryptography to authenticate entities
—whether that entity is a server process or a human user.
When messages are exchanged between entities, there is a risk that a message can be intercepted and that a process or a user can become impersonated. Certificate authorities
(CAs) are trusted parties that provide confidence that a public key truly belongs to an entity.
You request that the CA issue a certificate that contains your public key and that the CA has digitally signed the certificate. This is referred to as a signed certificate. The entity that receives your message acknowledges that your signed certificate is issued by a CA that it recognizes. If it recognizes the CA, the entity is able to substantiate your identity.
Figure 13.1 Types of Certificates
What Are Certificates?
187
Establishing trust is accomplished through certificates issued by three types of CAs:
• self-signed
You serve as your own CA and obtain a certificate using the certificate management tool on your machine.
• site-signed
You go to the IT department at your site to obtain a certificate.
• third-party-signed
You go to a commercial third-party CA, such as Comodo, to purchase a certificate.
For more information about TLS, see “Providers of Encryption” in Encryption in SAS .
188
Chapter 13 • Encryption Model
How SAS Uses Certificates
SAS Components That Use Certificates
SAS Deployment Agent
SAS Deployment Agent and its remote client perform configuration management operations for clustering and backups. There is one SAS Deployment Agent and its remote client per machine in the deployment. SAS Deployment Agent uses a Java keystore to store its secure certificate.
SAS Deployment Wizard configures communication between SAS Deployment Agent and its remote client using two-way client authentication with auto-generated self-signed certificates by default. Therefore, the deployment agent and its remote client each has to have access to a keystore and truststore. When multiple machines are involved, having a single keystore and truststore is necessary so that all of the remote clients can communicate with all of the deployment agents.
Credentials are generated during the deployment of the SAS Metadata Server. These credentials are retrieved from the metadata server during the configuration of other SAS servers. If there are multiple metadata servers in a deployment, there is a mechanism in
SAS Deployment Manager that enables you to upload a set of credentials to additional metadata servers. For more information, see “Manage SAS Deployment Agent Service” in the SAS Deployment Wizard and SAS Deployment Manager: User’s Guide .
We recommend that users who are unfamiliar with security setup use the default selfsigned certificates. If your site requires a site-signed or third-party-signed certificate, then you must supply paths to both the deployment agent and remote client keystore and truststore locations and their respective passwords when prompted by the deployment
wizard. For more information, see “Providing Your Own Certificates” on page 215 .
T I P
If you want to provide your own certificates, we recommend that you initially configure the deployment agent with auto-generated self-signed certificates that SAS
Deployment Wizard provides. Later, you can replace the auto-generated self-signed certificates with certificates that you provide, which saves you several manual configuration steps.
SAS Environment Manager
SAS Environment Manager is an operational monitoring and management system for
SAS deployments. SAS Environment Manager incorporates Hyperic technology from
VMware to offer enterprise-class operational features.
During installation, SAS Deployment Wizard will automatically create self-signed certificates for you. However, you can specify your own site-signed or third-party-signed certificates instead. SAS Environment Manager uses a Java keystore to store its private key and server certificate.
SAS Web Server
SAS Web Server is an HTTP server based on Pivotal Web Server.
Unlike the other SAS middle-tier components, SAS Web Server is a native application built with OpenSSL and does not use Java keystores. SAS Web Server requires the following:
How SAS Uses Certificates
189
• A private key that is in RSA format, is not protected with a passphrase, and is saved in the ASCII (Base64-encoded) PEM format
• A certificate that contains the server’s public key in X.509 certificate form and is saved in ASCII (Base64-encoded) PEM format.
Note: There is a peculiarity with the SAS Web Server in how it communicates with
HTTPS clients. To address this peculiarity, if you are providing your own certificates, in addition to adding the CA root certificate, you must add all CA intermediate certificates to the trusted CA bundle (and the Windows certificates stores).
SAS Web Application Server
SAS Web Application Server is based on Pivotal tc Server. SAS Web Application Server uses a Java keystore to store its private key and server certificate.
Your deployment might include additional instances of the web application server. If your deployment includes a SAS solution, the web applications related to the solution might be deployed to managed servers with names like SASServer2_1 or
SASServer12_1.
SAS Deployment Wizard does not include an option to automate HTTPS configuration for SAS Web Application Server. To implement certificates for communication with
SAS Web Application Server, you must perform manual steps after running the deployment wizard. For more details, see “Configuring SAS Web Application Server to
Use HTTPS” in SAS Intelligence Platform: Middle-Tier Administration Guide .
The Trusted CA Bundle
The Mozilla bundle is the list of CA certificates that are distributed with Mozilla software products. (For more information, see https://wiki.mozilla.org/
CA:IncludedCAs .)
Starting with the third maintenance release for SAS 9.4, the deployment wizard copies the Mozilla bundle (cacerts.pem and cacerts.jks) to create the trusted CA bundle
(trustedcerts.pem and trustedcerts.jks). These files reside in
SAS-installation-
directory/SASSecurityCertificateFramework/1.1/cacerts
. On UNIX,
SAS processes that are not Java use the trustedcerts.pem file to validate certificates. On
Windows, SAS processes that are not Java use the Windows certificates stores to validate certificates.
The deployment wizard makes a copy of trustedcerts.jks called jssecacerts, and moves jssecacerts to
SAS-installation-directory/
SASPrivateJavaRuntimeEnvironment/9.4/jre/lib/security
. SAS Java processes use the jssecacerts file to validate certificates. If this file resides in the same folder with cacerts, jssecacerts has precedence.
To add or remove CA certificates to or from the trusted CA bundle, you must use SAS
Deployment Manager. For more information, see
“Managing Certificates in the SAS
Note: Prior to the third maintenance release for SAS 9.4, you must add your CA
certificates to the SAS Private JRE using the
keytool -importcert
command.
For more information, see “Add Your Certificates to the SAS Private JRE” in
Encryption in SAS
.
190
Chapter 13 • Encryption Model
Figure 13.2 How SAS Deployment Wizard Installs the Trusted CA Bundle
Managing Certificates in the SAS Environment
Starting with the third maintenance release for SAS 9.4, SAS Deployment Manager has been enhanced with a functionality that enables you to add or remove CA root and intermediate certificates to or from the trusted CA bundle.
Note: Prior to the third maintenance release for SAS 9.4, you must add your CA
certificates to the SAS Private JRE using the
keytool -importcert
command.
For more information, see “Add Your Certificates to the SAS Private JRE” in
Encryption in SAS
.
CAUTION:
It is important to always perform the trusted CA bundle tasks in SAS
Deployment Manager when adding or removing certificates to or from your SAS
How SAS Uses Certificates
191 environment. Failing to perform these tasks in the deployment manager leads to security failures.
The following figure shows the process for adding certificates to the trusted CA bundle:
Figure 13.3 Adding Certificates to the Trusted CA Bundle
If you are providing your own certificates, there is an order in which you must follow.
The order is important when you are first deploying your environment. First, install the
SAS middle tier (if it is not already installed). Second, add your CA-signed root certificate and intermediate certificates to the trusted CA bundle. Then, configure the
SAS middle tier. Your certificates must be in the trusted CA bundle when SAS
Deployment Wizard loads specific products into SAS Content Server during the configuration phase of deployment. (The wizard’s load content step acts as a client and sends requests through SAS Web Server. This step fails if SAS Web Server uses the trusted CA bundle and cannot establish trust of your self-signed or site-signed certificates.)
Web applications running in SAS Web Application Server act as clients of other web applications and direct requests to web applications on other instances of SAS Web
Application Server. These requests are directed through SAS Web Server. Thus, each
SAS Web Application Server instance must trust the certificates used by SAS Web
Server. This trust is established by adding CA certificates to the trusted CA bundle.
For more information, see “Add Your Certificates to the Trusted CA Bundle” in
Encryption in SAS
.
192
Chapter 13 • Encryption Model
On Windows middle-tier machines only, in addition to adding your CA certificates to the trusted CA bundle, you must add your CA certificates to the Windows certificates stores using the Windows Certificates Snap-in.
For more information, see “Add Your Certificates to the Windows CA Stores” in
Encryption in SAS
.
Certificate Formats Supported by SAS Middle-Tier Components
The following table lists the SAS middle-tier components and the TLS certificate formats that they support:
Table 13.3 TLS Certificate Formats for SAS Middle-Tier Components
SAS Middle-Tier
Component
SAS Deployment Agent
SAS Environment Manager
SAS Web Server
Certificate Format
Supported
Java keystore
(One JKS file)
Java keystore
(One JKS file)
RSA private key
X.509 certificate
(Two PEM files)
SAS Web Application Server Java keystore
(One JKS file)
CA Root and Intermediate
Certificates Location
SAS Private JRE (jssecacerts)
SAS Private JRE (jssecacerts)
Windows: Windows certificates stores
SAS Private JRE (jssecacerts)
Obtaining Certificates
Overview of Obtaining Signed Certificates
The following SAS middle-tier components require one file—a Java keystore:
• SAS Deployment Agent
• SAS Environment Manager
• SAS Web Application Server
A Java keystore contains the following:
• private key
• CA root certificate
• CA intermediate certificates (if any)
• CA-signed certificate for the machine
Obtaining Certificates
193
The following figure summarizes the steps for obtaining a site-signed or third-partysigned certificate for a SAS middle-tier component (other than SAS Web Server):
Figure 13.4 Steps for Obtaining a Site-Signed or Third-Party-Signed Certificate for a SAS Middle-Tier Component
SAS Web Server requires two files:
• private key
This private key is in RSA format, is not protected with a passphrase, and is saved in the ASCII (Base64-encoded) PEM format.
• CA-signed certificate
This certificate contains the server’s public key in X.509 certificate form and is saved in ASCII (Base64-encoded) PEM format.
The following figure summarizes the steps for obtaining a site-signed or third-partysigned certificate for SAS Web Server:
194
Chapter 13 • Encryption Model
Figure 13.5 Steps for Obtaining a Site-Signed or Third-Party-Signed Certificate for SAS Web Server
How Many Private Keys and Signed Certificates Do I Need?
If you plan to provide your own signed certificates, you need to know how many to provide.
Each middle-tier machine requires at least one private key and one signed certificate to share between the SAS middle-tier components.
Except for SAS Web Server, the SAS middle-tier components require a private key and signed certificate in a Java keystore. SAS Web Server requires a private key and signed certificate in PEM format. For this reason, if SAS Web Server is located on the same machine with one or more SAS middle-tier components, you might find it easier to create a separate private key and signed certificate for SAS Web Server.
Note: If you want to use only one private key and signed certificate per machine, you
must convert the signed certificate to the proper format. For more information, see https://docs.oracle.com/cd/E35976_01/server.740/es_admin/src/ tadm_ssl_convert_pem_to_jks.html
.
For example, if you are implementing the middle tier on machine A and the server tier on machine B, then you would require a total of three signed certificates:
1cert_machA.jks, 2cert_machB.jks, and 3cert_machB.pem.
Table 13.4 A List of Machines and Their Certificates in a Sample SAS Middle-Tier
Deployment
SAS Components
SAS Metadata Server
Component
Location
Machine A
SAS Deployment Agent Machine A
SAS Environment Manager Agent Machine A
SAS Server Tier Machine B
SAS Deployment Agent Machine B
Signed Certificate File
Not applicable
1cert_machA.jks
1cert_machA.jks
Not applicable
2cert_machB.jks
SAS Components
SAS Environment Manager
Server
SAS Web Application Server
SAS Web Server
Implementing Certificates
195
Component
Location
Machine B
Signed Certificate File
2cert_machB.jks
Machine B
Machine B
2cert_machB.jks
3cert_machB.pem
Implementing Certificates
The options for implementing certificates consist of the following:
• self-signed certificates automatically generated by SAS Deployment Wizard
• certificates that you provide
• combination of automatically generated certificates and provided certificates
The following table lists the certificate options for the SAS middle-tier components:
Table 13.5 Certificate Options for the SAS Middle-Tier Components
Middle-Tier
Component
SAS Deployment
Agent
SAS Environment
Manager
Certificate Source
Auto-
Generated You Provide Combination
Self-signed only
Any type
*
Self-signed
**
or
Any type
Self-signed only
Any type
None Any type
Self-signed or
Any type
Any type
**
SAS Web Server
SAS Web Application
Server
None Any type Any type
*
**
All instances of “Any type” refers to self-signed, site-signed, and third-party-signed certificates that you provide.
Self-signed certificate auto-generated by SAS Deployment Wizard.
For more information, see
“Implementing Certificates” on page 213
.
196
Chapter 13 • Encryption Model
197
Chapter 14
Encryption Tasks
How to Maximize Encryption of Stored Passwords . . . . . . . . . . . . . . . . . . . . . . . .
Set or Change a Passphrase (SAS Deployment Manager) . . . . . . . . . . . . . . . . . . . 199
Set or Change a Passphrase (METADATA Procedure) . . . . . . . . . . . . . . . . . . . . . 199
How Password Re-encryption Activity Is Logged . . . . . . . . . . . . . . . . . . . . . . . . . 202
How to Change Encoding for Data in Motion for IOM Servers . . . . . . . . . . . . . .
How to Change Encoding for Data in Motion for Outbound Passwords . . . . . . .
How to Configure TLS between the Metadata Server and an LDAP Server . . . .
Obtaining a Signed Certificate for SAS Web Server . . . . . . . . . . . . . . . . . . . . . . . 210
Obtaining a Signed Certificate for Middle-Tier Components . . . . . . . . . . . . . . . . 211
Using a Combination of Auto-Generated and Your Own Certificates . . . . . . . . . . 219
How to Enable IOM Communication over HTTPS Proxy . . . . . . . . . . . . . . . . . . .
Connecting to an IOM Server Via Proxy While Using an Environment Variable . 221
Registering Proxy Information in SAS Metadata Server For Other Servers . . . . . 221
198
Chapter 14 • Encryption Tasks
How to Maximize Encryption of Stored Passwords
Passwords in Login Objects
To use the strongest available encryption for passwords that are stored in metadata:
1. Configure the metadata server to store any new or updated passwords using AES with 64-bit salt (SAS004).
a. Locate your equivalent of
SAS/Config/Lev1/SASMeta/
MetadataServer/omaconfig.xml
and open the file for edit.
T I P
it.
It is a good practice to make a backup copy of the file before you modify b. Set
STOREPASSWORDS="SAS004"
.
c. Restart the metadata server.
2. (Optional) Convert currently stored passwords to the SAS004 format. See “How to
Re-Encrypt Stored Passwords” on page 199 .
3. (Optional) Supply a passphrase to use as the encryption key, instead of using the default fixed key. See
“How to Re-Encrypt Stored Passwords” on page 199
.
Passwords for Internal Accounts
To use SHA-256 hashing for internal account passwords, set the HashPasswords option in the metadata server’s omaconfig.xml file. See
“How to Change Internal Account
Passwords in Configuration Files
To use the strongest available encryption for passwords that are stored in configuration files:
1. Use the PWENCODE procedure to prepare an AES with 64-bit salt (SAS004) version of each managed password.
2. Use the SAS Deployment Manager's password update utility to update the deployment with the SAS004-encoded version of the managed passwords. See
“Update a Managed Password” on page 35
.
See Also
• “Reference Information for omaconfig.xml” in SAS Intelligence Platform: System
Administration Guide
• “PWENCODE” in Encryption in SAS
How to Re-Encrypt Stored Passwords
199
How to Re-Encrypt Stored Passwords
Introduction
An unrestricted user can use the SAS Deployment Manager or the METADATA procedure to change how stored passwords are encrypted. These methods provide a thorough, secure, and efficient approach for an unrestricted user to address situations such as the following:
• For greater security, you want to re-encrypt stored passwords using a passphrase as the encryption key, instead of using the default fixed key.
• You suspect one or more stored passwords have been compromised, so you want to re-encrypt using a new passphrase.
• You have upgraded the encryption type with which new or updated passwords are stored, and you also want to upgrade the format of currently stored passwords.
• You want to downgrade the format of currently stored passwords.
CAUTION:
When you use the following instructions to introduce, change, or clear a passphrase, all passwords in login objects within the specified metadata server are affected.
Make sure you have a current backup before you set a passphrase.
Note: Other passphrases (for example, the passphrase for the SAS Web Server) are
distinct from the passphrases that are described in this topic and are not affected by these instructions.
Set or Change a Passphrase (SAS Deployment Manager)
To change a passphrase using the SAS Deployment Manager:
1. On the metadata server's host, navigate to your equivalent of
SAS-installation-
directory/SASDeploymentManager/9.4/
and launch sasdm.exe (Windows), sasdm.sh (UNIX), or sasdm.rexx (z/OS).
Windows Specifics
On Windows, log on as a Windows administrator of the current machine.
Operating Environment Information
On UNIX and z/OS, log on with the user ID that was used to perform the SAS installation.
2. In the SAS Deployment Manager, select the change passphrase task and log on as an unrestricted user.
3. Change the passphrase. For assistance with the user interface, see the Help within the application.
Set or Change a Passphrase (METADATA Procedure)
To use a passphrase as the encryption key for passwords that are stored in metadata:
200
Chapter 14 • Encryption Tasks
1. Verify that the metadata server is configured to store passwords in SAS004 or
SAS003 format. Passphrase keys are supported for only these formats.
Note: The format in which the metadata server stores passwords is controlled by the
STOREPASSWORDS="SAS00n" setting in your equivalent of
SAS/Config/
Lev1/SASMeta/MetadataServer/omaconfig.xml
. Changes to this file take effect when you restart the metadata server.
2. In a SAS session, connect to the metadata server as an unrestricted user and pause the server for administration as follows.
options
metaserver='machine.company.com'
metaport=8651
metauser='sasadm@saspw'
metapass='secret';
proc metaoperate
action=PAUSE
options='<Server STATE="ADMIN"/>';
run;
3. In the SAS session, set a passphrase as follows. proc metadata
in="<ChangePassPhrase>
<PassPhrase>My passphrase</PassPhrase>
</ChangePassPhrase>";
run;
The preceding code decrypts currently stored passwords and then re-encrypts them using the passphrase that you supply.
T I P
You can use the PWENCODE procedure to prepare an encrypted version of your passphrase, and supply the encrypted version of the passphrase in the preceding code.
4. To verify results, inspect the metadata server log. See “How Password Re-encryption
Activity Is Logged” on page 202
.
5. In the SAS session, resume the metadata server.
proc metaoperate
action=RESUME;
run;
To re-encrypt stored passwords using a new passphrase, follow the preceding instructions and supply the new passphrase in step 3 as follows: proc metadata
in="<ChangePassPhrase>
<PassPhrase>My new passphrase</PassPhrase>
</ChangePassPhrase>";
run;
To remove a passphrase (and use the default fixed key instead), follow the preceding instructions and submit a blank passphrase in step 3 as follows: proc metadata
in="<ChangePassPhrase>
<PassPhrase/>
</ChangePassPhrase>";
run;
How to Re-Encrypt Stored Passwords
201
If you must maintain a few stored passwords in SAS002 format, see
.
Downgrade Stored Passwords
To downgrade the encryption type of stored passwords:
1. Configure the metadata server to store any new or updated passwords in the downgraded format. a. Locate your equivalent of
SAS/Config/Lev1/SASMeta/
MetadataServer/omaconfig.xml
and open the file for edit.
T I P
It is a good practice to make a copy of the file before you modify it.
b. Set the following options:
•
STOREPASSWORDS="SAS002"
•
HashPasswords="MD5"
Note: The latter setting is within the
InternalAuthenticationPolicy
section of the omaconfig.xml file. It affects passwords in internal accounts.
For details, see “How to Change Internal Account Policies” on page 177 .
c. Restart the metadata server.
2. In a SAS session, connect to the metadata server as an unrestricted user and pause the server for administration as follows.
options
metaserver='machine.company.com'
metaport=8651
metauser='sasadm@saspw'
metapass='secret';
proc metaoperate
action=PAUSE
options='<Server STATE="ADMIN"/>';
run;
3. In your SAS session, set a blank pass phrase as follows: proc metadata
in="<ChangePassPhrase>
<PassPhrase/>
</ChangePassPhrase>";
run;
Notice that the
PassPhrase
element is empty. The preceding code decrypts stored passwords and then stores them in a downgraded format.
4. To verify results, inspect the metadata server log. See “How Password Re-encryption
Activity Is Logged” on page 202
.
5. Resume the metadata server as follows: proc metaoperate
action=RESUME;
run;
6. Update the password in each internal account as follows:
•
.
202
Chapter 14 • Encryption Tasks
• For other internal accounts, use SAS Management Console. See “Reset an
Internal Password” in SAS Management Console: Guide to Users and
Permissions
.
How Password Re-encryption Activity Is Logged
As the passwords within each repository are converted, repository-level messages are written to the metadata server log. Here are some examples of the messages:
ChangePassPhrase: repository Id=A5KUG99E, Name=repos1, PauseState=ADMIN is being processed.
ChangePassPhrase: repository Id=A5KUG98E, Name=repos2 is not being processed because the
PauseState=OFFLINE must be ADMIN.
To log changes to individual objects, add the following content to the metadata server’s log4sas.xml file:
<logger name="Audit.Meta.Updates.PublicObjects">
<level value="Trace"/>
</logger>
<logger name="Audit.Meta.Updates.Projects.PublicObjects">
<level value="Debug"/>
</logger>
Here are some examples of the log messages:
Audit Public Object Type=Login Name=user2 ObjId=A5KUG99E.AQ000001 has been updated.
Audit Public Object Type=Login Name=user3 ObjId=A5KUG99E.AQ000002 has been updated.
Audit Public Object Type=Login Name=sasadm ObjId=A5KUG99E.AQ000003 has been updated.
Note: The log messages are written after the objects are committed to the repository.
Changes to SASPassword objects are not logged.
See Also
• “Connection Options ” in SAS Language Interfaces to Metadata
• “METAOPERATE” in SAS Language Interfaces to Metadata
• “About Metadata Server Loggers” in SAS Intelligence Platform: System
Administration Guide
• “Reference Information for omaconfig.xml” in SAS Intelligence Platform: System
Administration Guide
• “PWENCODE” in Encryption in SAS
How to Change Encoding for Data in Motion for
IOM Servers
Automatic Configuration
When you install the metadata server, you select an encryption level (which traffic content is encrypted) and an encryption algorithm (how that traffic is encrypted). In
How to Change Encoding for Data in Motion for IOM Servers
203
order to connect to the metadata server, clients must conform to the metadata server’s encryption requirements.
Note: If the FIPS encryption algorithm was set during the initial deployment, and you
manually change the server encryption algorithm, you must also remove the -
encryptfips option from the following files:
• ObjectSpawner.bat (Windows) or ObjectSpawner.sh (UNIX)
• ConnectSpawner.bat (Windows) or ConnectSpawner.sh (UNIX)
For more information about the encryptfips option, see “ENCRYPTFIPS System
Option” in Encryption in SAS .
Instructions
If you need to change encryption settings for data in motion after installation is complete, use the following instructions.
1. Update server configuration files as follows: a. In the operating system that hosts the metadata server, navigate to your equivalent of
SAS/Config/Lev1/SASMeta/MetadataServer/
.
• To change the algorithm, add the NETENCRALG setting that you need to the sasv9_usermods.cfg file.
• To change the encryption level, copy the entire OBJECTSERVERPARMS line from the sasv9.cfg file into the sasv9_usermods.cfg file. Then edit the
CEL value in the usermods version of the file.
For example, to encrypt all traffic with AES, add these lines:
-netencralg "AES"
-objectserverparms "cel=everything {other-parameters}"
On z/OS, exclude the initial hyphens and add equal signs as follows: netencralg="AES" objectserverparms="cel=everything {other-parameters}" b. (Optional) If your deployment offers direct connections from clients to the OLAP server, make the same updates to that server's configuration file.
Note: The OLAP server's configuration file is in your equivalent of
SAS/Config/Lev1/SASApp/OLAPServer/
.
2. Update server metadata definitions as follows: a. In SAS Management Console, under Server Manager, select the metadata server's definition .
Note: To get to the server definition, you must expand the application server
node and the logical server node .
b. Right-click the first connection object , and select Properties. c. In the Connection dialog box, select the Options tab and click Advanced
Options. Adjust the settings as necessary.
d. In the Advanced Options dialog box, select the Encryption tab.
Repeat the preceding steps for each server that is launched by the object spawner
(the stored process server, the workspace server, and the pooled workspace server).
204
Chapter 14 • Encryption Tasks
3. Stop, restart, and validate the servers.
About NETENCRALG and CEL
On direct connections, encryption is governed by the server's invocation command. Here are details and some examples:
NETENCRALG (network encryption algorithm) is a SAS system option. The NETENCRALG setting that is defined for the metadata server during installation is in the metadata server's sasv9.cfg file.
• If you accept the default encryption settings during installation, the configuration file includes this line:
-netencralg "SASProprietary"
• If you selected the AES algorithm during installation, the setting in the metadata server's sasv9.cfg file is as follows:
-netencralg "AES"
• If a different NETENCRALG setting has been added to the metadata server's sasv9_usermods.cfg file, that setting has priority.
• Other supported values for NETENCRALG are DES, TripleDES, RC4, and RC2.
CEL (client encryption level) is a parameter in the OBJECTSERVERPARMS SAS system option. The CEL setting that is defined for the metadata server during installation is in the metadata server's sasv9.cfg file.
• If you accept the default encryption settings during installation, the configuration file includes this line:
-objectserverparms "cel=credentials {other-parameters}"
• If, during installation, you selected the option to encrypt all traffic, the setting in the metadata server's sasv9.cfg file is as follows:
-objectserverparms "cel=everything {other-parameters}"
• If a different CEL setting has been added to the metadata server's sasv9_usermods.cfg file, that setting has priority.
Note: On z/OS, the preceding syntax examples are slightly different. See step 1a in the
preceding topic.
It is not necessary to specify encryption settings in the invocation command for every component for the following reasons:
• Encryption algorithm and level are negotiated between each pair of communicating components. For example, when the OLAP server and object spawner initialize, they contact the metadata server and conform to the metadata server's encryption settings.
The same negotiation occurs when a client application contacts a server.
• For spawned servers (the stored process server, the pooled workspace server, and the workspace server), encryption is determined by metadata settings, not by a server invocation command.
See Also
“Encryption for Data in Motion” on page 185
How to Change Encoding for Data in Motion for Outbound Passwords
205
How to Change Encoding for Data in Motion for
Outbound Passwords
Outbound Passwords
A password is outbound when a client retrieves the password from the metadata in order to provide seamless access to a server such as Oracle. The password is outbound from the perspective of the metadata server. Connections to third-party servers often use outbound passwords. Most other connections do not use outbound passwords.
Instructions
In the initial configuration, outbound passwords are transmitted in SAS002 format
(SASProprietary).You can choose to increase the encoding strength for outbound passwords to AES 256 with 16-bit salt (SAS003) or AES 256 with 64-bit salt (SAS004).
To increase encryption strength for outbound passwords:
1. Edit the metadata server's omaconfig.xml file to change the initial setting,
RETURNPASSWORDS="SAS002", to a more secure setting (for example,
RETURNPASSWORDS="SAS004"). The metadata server's omaconfig.xml file is located in your equivalent of
SAS/Config/Lev1/SASMeta/
MetadataServer/
.
Note: For more information, see the RETURNPASSWORDS option in
“Reference
Information for omaconfig.xml” in SAS Intelligence Platform: System
Administration Guide
.
2. Restart the metadata server.
3. Verify that server connections continue to function as expected. If you encounter problems, either review the following topic or revert to
RETURNPASSWORDS="SAS002".
Compatibility
Almost all connections involve a SAS server, and SAS servers can decode SAS004 passwords. For example, connections from SAS Information Map Studio to an Oracle server go through a workspace server. The workspace server decodes the outbound
Oracle password.
However, a few specialized connections run directly from a Java client or .NET client to a third-party server. These clients cannot decode SAS003 and SAS004 passwords. This is a deliberate limitation that reduces security exposures. Of course, a third-party server cannot decode SAS003 and SAS004 passwords either. As a result, such connections fail if they attempt to use a password that is in either of those formats. Here are some examples of connections that cannot use a SAS003 or SAS004 password:
• Connections from a Java client or .NET client to an Esri server.
Note: The Esri server uses host authentication. If possible, avoid the use of
outbound passwords by locating this server on a machine that recognizes the
206
Chapter 14 • Encryption Tasks
accounts with which users log on to SAS applications. This facilitates use of cached credentials to access the Esri server.
• Certain connections within the following solutions:
• SAS Profitability Management
• SAS Activity-Based Management
• SAS Merchandise Intelligence
• SAS Model Manager (in some configurations)
• SAS LASR Analytic Server
• Internal database management system (DBMS) accounts (for example, the administrator of the SAS Web Infrastructure Platform Data Server, dbmsowner).
If your deployment requires connections that are incompatible with SAS003 or SAS004 passwords, choose either of the following approaches:
• Preserve the initial setting of RETURNPASSWORDS="SAS002".
Note: With the STOREPASSWORDS="SAS003" and
RETURNPASSWORDS="SAS002" settings, outbound passwords are stored in
SAS003 format and downgraded to SAS002 format before they are transmitted.
• Set RETURNPASSWORDS="SAS003", but also selectively force the outbound passwords for the problematic connections to be transmitted in SAS002 format. To force a particular password to be transmitted in SAS002 format, store that password in that format. A password that is stored in SAS002 format is transmitted in that format even if RETURNPASSWORDS="SAS003", because the metadata server does not upgrade the encoding strength of a stored password when the password is transmitted.
To use this approach:
1. In the metadata server's omaconfig.xml file, set
STOREPASSWORDS="SAS002". Restart the metadata server.
2. In SAS Management Console, under User Manager, locate and select the login that contains the outbound password. The login is on the Accounts tab of a user or group definition.
3. Click Edit. In the Login Properties dialog box, enter and confirm the password.
Click OK to close the Login Properties dialog box. Click OK again to close the user or group properties dialog box.
Note: This stores the password in the metadata in SAS002 format.
4. Repeat steps 2 and 3 for any other problematic outbound passwords.
5. In the metadata server's omaconfig.xml file, upgrade the settings (for example, set STOREPASSWORDS="SAS004" and RETURNPASSWORDS="SAS004").
Restart the metadata server.
6. Verify that server connections continue to function as expected.
Note: Anytime in the future that you update the SAS002 passwords, you must repeat
the preceding steps process in order to store the new password in SAS002 format.
How to Configure TLS between the Metadata Server and an LDAP Server
207
How to Configure TLS between the Metadata
Server and an LDAP Server
Before You Begin
This is not a universally necessary task. The following instructions support encryption of direct communication between the metadata server and an LDAP server using the
Transport Layer Security (TLS) protocol. This content is also applicable to the predecessor protocol, Secure Sockets Layer (SSL).
This functionality applies to only specialized configurations in which the metadata
server directly uses an LDAP server as an authentication provider. See “Direct LDAP
.
Configuration Instructions
1. On the LDAP server: a. Make sure that a port is configured for TLS. b. Examine the Client Certification setting and note its implications for your work in step 2e below.
• If the setting specifies that a client certificate is optional, step 2e below is optional for you. This is the default setting on most LDAP servers.
• If the setting specifies that a client certificate is required, step 2e below is mandatory for you.
• If the setting specifies that a client certificate is not allowed, a client certificate is ignored, so there is no reason for you to perform step 2e below.
2. On the metadata server: a. If support is not already in place for direct LDAP, configure that support. See
“How to Configure Direct LDAP Authentication” on page 160
.
b. Make sure that the value that is specified in the LDAP_PORT (or AD_PORT) environment variable matches a TLS-configured port on the LDAP server. c. In the same location where you set environment variables to configure the metadata server for direct LDAP, add the LDAP_TLSMODE (or
AD_TLSMODE) environment variable, and set it to
1
. Setting this variable causes the metadata server to attempt to use TLS.
d. Provide a certificate authority (CA) certificate to use TLS.
Host
Windows
Details
The CA certificate is installed using Microsoft Certificate Services.
208
Chapter 14 • Encryption Tasks
Host
UNIX z/OS
Details
For the second maintenance release of SAS 9.4 and previous releases, the CA certificate file is provided in an accessible directory and referenced by the SSLCALISTLOC option.
*
In the third maintenance release of SAS 9.4, certificates are located in the trusted CA bundle (trustedcerts.pem) by default, and the
SSLCALISTLOC system option that points to this certificate file is automatically added at installation to the sasv9.cfg file. The sasv9.cfg file is located in
SAS-installation-directory/
SASFoundation/9.4
. For more information, see SAS 9.4
Intelligence Platform: Installation and Configuration Guide.
The CA certificate file is provided in an accessible directory and referenced by the SSLCALISTLOC option.
*
*
Add the option to the metadata server's invocation command (in the same location where you specified -authpd). For syntax, see "SAS System Options for Encryption" in Encryption in SAS.
e. If the LDAP server requires client certification (see step 1b above), specify the location of client certificates.
Host
Windows
UNIX z/OS
Details
The client certificate is installed using Microsoft Certificate
Services. The SSLCERTSUBJ, SSLCERTSERIAL, and
SSLCERTISS options are used to locate the client certificate.
*
The client certificate files are provided in an accessible directory.
The SSLCERTLOC, SSLPVTKEYLOC, and SSLPVTKEYPASS options can be used to locate the client certificate. PKCS12 certificates can also be used with the SSLPKCS12LOC and
SSLPKCS12PASS options.
*
The client certificate files are provided in an accessible directory.
The SSLCERTLOC, SSLPVTKEYLOC, and SSLPVTKEYPASS options can be used to locate the client certificate. PKCS12 certificates can also be used with the SSLPKCS12LOC and
SSLPKCS12PASS options.
*
*
Add the necessary options to the metadata server's invocation command (in the same location where you specified -authpd). For syntax, see "SAS System Options for Encryption" in
Encryption in SAS.
3. Restart the metadata server.
4. To test the results, use the following code. Replace the sample values with your metadata server connection information, and submit the code in a SAS code editor.
proc metaoperate
server="a123.us.company.com"
port=8561
userid="myLDAPuserID@myldap"
password="myLDAPpassword"
action=status; run;
How to Configure TLS between the Metadata Server and an LDAP Server
209
Connecting to an LDAP Load Balancer
Potential Problem
In the default TLS configuration, each connection includes a server name check to verify that the host name in the certificate of the node to which the metadata server connected matches the host name of the target LDAP server.
If the connection is to an LDAP server load balancer (rather than directly to the authenticating LDAP server), the server name check might fail, because the LDAP load balancer might be on a different node than the LDAP server to which the metadata server is redirected.
Preferred Solution: Wildcard Certificate
If you encounter the problem that is described in the preceding section, use a wildcard when specifying the subject name in the certificate.
For example, instead of specifying
myHost.myCompany.com
, specify
*.myCompany.com
.
CAUTION:
If you use a wildcard, only the domain portion of the server name in the certificate is validated against the server that you connect to.
Do not use this approach unless you encounter the problem that is described in the preceding section.
Note: Using a wildcard is the preferred solution because it provides domain name
validation, which reduces the risk of Domain Name System (DNS) redirection or hijacking.
Alternate Solution: Environment Variable
As an alternative to using a wildcard certificate, you can disable the server name check by defining an additional environment variable (NOSSLNAMECHECK=1) for the metadata server. For example, you might add the following setting to the sasv9.cfg file:
-set NOSSLNAMECHECK 1
Or, you can define the environment variable in the metadata server’s start-up script as follows: export NOSSLNAMECHECK=1
CAUTION:
In this approach, neither the server name nor the domain name is validated.
For this reason, using a wildcard certificate is the preferred solution to the problem that is described in the preceding section.
SASSSLREQCERT
Windows Specifics
Windows clients always request and verify server certificates. The
SASSSLREQCERT environment variable is supported only on UNIX.
The SASSSLREQCERT environment variable specifies what checks to perform on server certificates in a TLS session. This environment variable supports the following settings:
210
Chapter 14 • Encryption Tasks
DEMAND
A server certificate is requested. If no valid certificate is provided, the session terminates. DEMAND is the default setting.
TRY
The client requests a server certificate. If no certificate is provided, the session proceeds normally. If an invalid certificate is provided, the session terminates.
ALLOW
The client requests a server certificate, but the session proceeds normally even if no certificate is provided or an invalid certificate is provided.
NEVER
The client does not request a server certificate.
For example, add the follow setting to the script: export SASSSLREQCERT=ALLOW
See Also
•
“About Setting Environment Variables” on page 165
•
Encryption in SAS
How to Implement Certificates
Obtaining a Signed Certificate for SAS Web Server
If you have not used certificates in the past, you might want to obtain a site-signed or third-party-signed certificate. Although each SAS machine needs its own private key and server certificate, you can generate the private key and a certificate signing request
(CSR) on any machine. The most important thing is to safeguard the private key. This topic provides you with an overview of how to obtain a signed certificate for SAS Web
Server.
Unlike the other SAS middle-tier components that require their certificates in a Java keystore, SAS Web Server requires two files:
• private key
This private key is in RSA format, is not protected with a passphrase, and is saved in
ASCII (Base64-encoded) PEM format.
• CA-signed certificate
This certificate contains the server’s public key in X.509 certificate form and is saved in ASCII (Base64-encoded) PEM format.
This topic explains how to obtain a signed certificate for SAS Web Server on UNIX. For steps on Windows, see “TLS on Windows: Setting Up Digital Certificates ” in
Encryption in SAS
. To obtain a site-signed or third-party-signed certificate for SAS Web
Server, perform these steps:
1. Decide which type of CA to use at your site:
• site-signed
• third-party-signed
How to Implement Certificates
211
2. Establish a location on your machine where you will store your private key and signed certificates.
For the examples in this document,
/opt/certs
is used to store the private key and certificates.
3. Change the directory to where your
openssl
commands reside.
For example:
cd /usr/bin
4. Use the
openssl
command to create a private key.
For example:
./openssl genrsa -out /opt/certs/machine01_key.pem 2048
T I P
For more information about
openssl
commands, see https:// www.openssl.org/docs/manmaster/apps/openssl.html
.
5. Use the
openssl
command to create a CSR.
For example:
./openssl req -new -key /opt/certs/machine01_key.pem out /opt/certs/machine01_csr.pem
6. Submit your CSR file and your private key to your CA.
You should receive the following from your CA:
• signed certificate (containing the CA’s public key)
• CA root certificate
• CA intermediate certificate
7. Store the private key and certificates that you received from your CA in the location
that you established in Step 2
.
For more information, go to the link appropriate for your operating system.
• On Windows, see “TLS on Windows: Setting Up Digital Certificates ” in Encryption
in SAS
.
• On UNIX, see “ Setting Up Digital Certificates Using OpenSSL ” in Encryption in
SAS
.
• On z/OS, see “TLS on z/OS: Setting Up Digital Certificates” in Encryption in SAS .
Obtaining a Signed Certificate for Middle-Tier Components
If you have not used certificates in the past, you might want to obtain a site-signed or third-party-signed certificate. This topic provides you with an overview of how to obtain a signed certificate for a SAS middle-tier component (other than SAS Web Server).
Unlike SAS Web Server, SAS Web Application Server, SAS Environment Manager, and
SAS Deployment Agent require their certificates in a Java keystore.
Note: For commands to obtain signed certificates for SAS Web Server, see
Signed Certificate for SAS Web Server” on page 210 .
To obtain a site-signed or third-party-signed certificate for SAS middle-tier components
(other than SAS Web Server), perform these steps:
1. Decide which type of CA to use at your site:
212
Chapter 14 • Encryption Tasks
• site-signed
• third-party-signed
2. Establish a location on your machine where you will store your private key and signed certificates.
For the examples in this document,
/opt/certs
is used to store the private key and certificates.
T I P
If any SAS middle-tier component resides on the same machine as SAS
Environment Manager, we recommend that you use the path and password that
SAS Environment Manager uses by default (for example,
../../config/ hyperic.keystore
for Hyperic). By doing so, you will not have to manually specify a new path and password later in the configuration.
3. Change the directory to where your
keytool
command resides.
For example:
cd /usr/java/jdk1.8.0_45/bin
4. Use the
keytool
command to create a keystore. (By default, a keystore contains a private key.)
For example:
./keytool -genkey -alias hq -keyalg RSA -keystore /opt/ certs/hyperic.keystore -keysize 2048
T I P
For more information about the
keytool
command, see http:// docs.oracle.com/javase/7/docs/technotes/tools/windows/keytool.html
.
5. Use the
keytool
command to create a CSR.
For example:
./keytool -certreq -alias hq -keystore /opt/certs/ hyperic.keystore -file /opt/certs/hyperic.csr
6. Submit your CSR file and your private key to your CA.
You should receive the following from your CA:
• signed certificate (containing the CA’s public key)
• CA root certificate
• CA intermediate certificate
7. Store the private key and certificates that you received from your CA in the location
that you established in Step 2
.
8. Use the
keytool
command to add the CA root certificate, CA intermediate certificate, and the signed certificate (in this order) to your keystore.
a. Add the CA root certificate to your keystore.
In this example, root_ca.cer is the CA root certificate and hyperic.keystore is the keystore file:
./keytool -importcert -file /opt/certs/root_ca.cer keystore /opt/certs/hyperic.keystore -trustcacerts -alias rootca
b. Add the CA intermediate certificate to your keystore.
In this example, int_ca.cer is the CA intermediate certificate:
How to Implement Certificates
213
./keytool -importcert -file /opt/certs/int_ca.cer keystore /opt/certs/hyperic.keystore -trustcacerts -alias intca
c. Add the signed certificate to your keystore.
In this example, myserver_signed.cer is the signed certificate:
./keytool -importcert -file /opt/certs/ myserver_signed.cer -keystore /opt/certs/hyperic.keystore
-trustcacerts -alias myserver
9. Verify that the certificates that you added to your keystore are present:
./keytool -list –keystore path-to-keystore-file
In this example, hyperic.keystore is the keystore file:
You should see something similar to the following (look for the aliases that you added):
Keystore type: JKS
Keystore provider: SUN
Your keystore contains 4 entries myserver, Oct 16, 2015, trustedCertEntry,
Certificate fingerprint (SHA1): DA:62:4A:D6:14:79:28:46:0C:45:1D:1B:F8:
5C:29:BE:69:79:9D:2E intca, Oct 16, 2015, trustedCertEntry,
Certificate fingerprint (SHA1): 12:4D:C9:88:CD:D5:F3:E9:9E:29:D8:AB:F1:
00:AD:93:60:61:11:45 hq, Oct 16, 2015, PrivateKeyEntry,
Certificate fingerprint (SHA1): BD:9C:D3:C1:61:D1:0C:93:5D:4F:6D:49:77:
B1:40:C2:60:96:67:C3 rootca, Oct 16, 2015, trustedCertEntry,
Certificate fingerprint (SHA1): 16:E4:1E:AC:F0:0C:1E:DB:30:3A:1A:C3:EF:
4C:D6:D5:15:38:2A:25
For more information, go to the link appropriate for your operating system.
• On Windows, see “TLS on Windows: Setting Up Digital Certificates ” in Encryption
in SAS
.
• On UNIX, see “ Setting Up Digital Certificates Using OpenSSL ” in Encryption in
SAS
.
• On z/OS, see “TLS on z/OS: Setting Up Digital Certificates” in Encryption in SAS .
Implementing Certificates
The following list summarizes the steps to implement the various certificate options:
• Using self-signed certificates automatically generated by SAS Deployment Wizard.
For more information, see “Using Auto-Generated Certificates” .
This option consists of installing and configuring the SAS middle tier using the deployment wizard.
•
Providing your own certificates. For more information, see “Providing Your Own
214
Chapter 14 • Encryption Tasks
This option consists of these high-level steps:
1. If you do not already have them, obtain your certificates and private keys. Store these certificates and private keys on any machine that has the SAS components that use them.
2. Install and configure your SAS metadata tier and SAS server tier.
(Use SAS Deployment Wizard.)
Note: When prompted, use the auto-generated certificates for SAS Deployment
Agent and its remote client. Later, you can replace the auto-generated certificates with certificates that you provide.
3. On both Windows and UNIX on the SAS metadata-tier and SAS server-tier machines, depending on your release of SAS 9.4, add your CA root certificate and intermediate certificates either to the trusted CA bundle or to the SAS Private
JRE.
(Use SAS Deployment Manager or the Java
keytool -importcert
command.)
4. On Windows on the SAS metadata-tier and SAS-server tier machines, you must also add your CA root certificate and intermediate certificates to the Windows certificates store using the Windows Certificates Snap-in. (Most third-partysigned CA certificates already exist in the Windows certificates store.)
5. Install your SAS middle tier.
(Use SAS Deployment Wizard.)
6. On both Windows and UNIX middle-tier machines, add your certificates to the trusted CA bundle.
(Use SAS Deployment Manager.)
7. On Windows middle-tier machines, you must also add any pre-existing selfsigned and site-signed certificates to the Windows certificates stores using the
Windows Certificates Snap-in. (Most third-party-signed CA certificates already exist in the Windows certificates stores.)
8. Configure your SAS middle tier.
(Use SAS Deployment Wizard.)
9. Perform manual, post-deployment steps.
Note: Implementing certificates for SAS Web Application Server consists of
manual, post-deployment steps.
• Using a combination of self-signed certificates automatically generated and providing your own certificates.For more information, see
Auto-Generated and Your Own Certificates” .
This option has the same steps as providing your own certificates. The only difference is that you do not replace the auto-generated certificates with your own certificates.
Using Auto-Generated Certificates
If you want SAS Deployment Wizard to generate self-signed certificates to secure SAS
Deployment Agent and its remote client and the SAS Environment Manager server and its agents, perform these steps:
How to Implement Certificates
215
Note: Using auto-generated certificates secures your SAS Deployment Agent and SAS
Environment Manager only. To secure all of the components in your SAS middle tier, you must provide your own certificates or use a combination of your own certificates and certificates auto-generated by SAS Deployment Wizard. For more
information, see “Implementing Certificates” on page 195 .
1. Run SAS Deployment Wizard. Make sure that both Install SAS Software and
Configure SAS Software are selected.
2. When you are prompted for the SAS Deployment Agent remote communication configuration, select Generate credentials to secure the connection.
3. When you are prompted for the configuration prompting level, select Custom.
4. When you are prompted, make sure that you select Generate a default JKS format
keystore from SAS Environment Manager.
5. When you are prompted, make sure that you select Establish secure
communication.
Providing Your Own Certificates
For SAS Deployment Agent and its remote client and for the SAS Environment Manager server and its agents, SAS Deployment Wizard provides auto-generated, self-signed certificates by default. Because these certificates have intermachine connections, are behind your firewall, and do not involve users, there is less of a reason to implement site-signed or third-party-signed certificates. However, if you must use your own certificates, you can.
This topic explains how to accept the default action—that is, let SAS Deployment
Wizard install auto-generated self-signed certificates for you. After the deployment wizard is finished, you can go back and replace the auto-generated, self-signed certificates with certificates that you provide, which saves you several manual configuration steps.
To provide your own certificates to secure your SAS middle tier, follow these steps:
1. If you do not already have them, obtain your certificates. For more information, see
“Obtaining Certificates” on page 192
.
2. Decide on a location on each machine in your SAS environment where you will store your certificates. Make sure that your certificates reside in these locations.
3. Use the following table to record keystore and truststore locations and passwords for
SAS Deployment Agent and its remote client.
SAS Deployment Wizard prompts you for these locations and passwords during SAS installation and configuration. The deployment wizard encodes the passwords that you provide into SAS proprietary format (SAS002).
Note: Your keystore must be in JKS format.
T I P
We recommend that instead of providing your own certificates to SAS
Deployment Agent during installation, that you accept the default and let SAS
Deployment Wizard install auto-generated self-signed certificates for you. After the deployment wizard is finished, you can go back and replace the autogenerated, self-signed certificates with certificates that you provide, which saves you several manual configuration steps.
216
Chapter 14 • Encryption Tasks
Table 14.1 SAS Deployment Agent Keystore and Truststore Information
Keystore and Truststore
Properties Actual Value You Are Using
SAS Deployment Agent keystore path
SAS Deployment Agent keystore password
SAS Deployment Agent truststore path
SAS Deployment Agent truststore password
SAS Deployment Agent remote client keystore path
SAS Deployment Agent remote client keystore password
SAS Deployment Agent remote client truststore path
SAS Deployment Agent remote client truststore password
4. Use the following table to record keystore locations and passwords for the SAS
Environment Manager server and its agents.
SAS Deployment Wizard prompts you for these locations and passwords during SAS installation and configuration. The deployment wizard encodes the passwords that you provide into SAS proprietary format (SAS002).
Table 14.2 SAS Environment Manager Keystore Information
Keystore Properties
SAS Environment Manager server keystore path
Actual Value You Are Using
SAS Environment Manager server keystore password
SAS Environment Manager alias
SAS Environment Manager agent keystore path
How to Implement Certificates
217
Keystore Properties
SAS Environment Manager agent keystore password
Actual Value You Are Using
5. Use the following table to record certificate and private key locations for SAS Web
Server. SAS Deployment Wizard prompts you for these locations and passwords during SAS installation and configuration.
Table 14.3 SAS Web Server Private Key and Certificate Information
Private Key and
Certificate Properties
SAS Web Server X.509 certificate path
SAS Web Server RSA private key path
Actual Value You Are Using
6. Use the following table to record keystore locations and passwords for SAS Web
Application Server and its agents. SAS Deployment Wizard does not prompt you for these locations and passwords. You must manually update the SAS Web Application
Server configuration with this information after you deploy SAS.
T I P
You can use the same keystore for SAS Environment Manager and SAS Web
Application Server if they reside on the same machine.
Table 14.4 SAS Web Application Server Keystore Information
Actual Value You Are Using Keystore Properties
SAS Web Application
Server keystore path
SAS Web Application
Server keystore password
7. Install and configure your SAS metadata-tier and SAS-server tier machines. For more information, see
“Using Auto-Generated Certificates” on page 214
.
8. Add your CA root certificate and intermediate certificates to your SAS metadata-tier and SAS server-tier machines. For more information, see “Overview of Certificate
Management Using the SAS Deployment Manager” in Encryption in SAS .
9. On each of your middle-tier machines, start SAS Deployment Wizard by navigating to
SAS-installation-directory/SASDeploymentManager/9.4
, and launch
setup.exe
(Windows) or
setup.sh
(UNIX). On Windows, you can use the shortcut on the Start menu.
10. On the Select Deployment Type page, select Perform a Planned Deployment and
Install SAS Software.
Note: Make sure that Configure SAS Software is deselected.
218
Chapter 14 • Encryption Tasks
11. When prompted by the deployment wizard, select Generate credentials to secure
the connection, and click Next.
T I P
We recommend that instead of providing your own certificates to SAS
Deployment Agent during installation, that you accept the default and let SAS
Deployment Wizard install auto-generated self-signed certificates for you. After the deployment wizard is finished, you can go back and replace the default SAS
Environment Manager auto-generated JKS keystore with certificates that you provide. Choosing Generate credentials to secure the connection saves you several manual configuration steps.
12. On the Additional Resources page, click Finish to close the deployment wizard.
13. If you have additional middle-tier machines, repeat steps 4 through 9 on each middle-tier machine.
Otherwise, proceed to the next step.
14. Perform one of the following steps based on your operating system:
• Windows
• Depending on your release of SAS 9.4, add your CA root certificate and intermediate certificates either to the trusted CA bundle or to the SAS Private
JRE on each of your middle-tier machines. For more information, see “Add
Your Certificates to the Trusted CA Bundle” in Encryption in SAS or “Add
Your Certificates to the SAS Private JRE” in Encryption in SAS .
• Using the Windows Certificates Snap-in, add your CA root certificate and intermediate certificates to the Windows certificates stores on each of your middle-tier machines. For more information, see “Add Your Certificates to the Windows CA Stores” in Encryption in SAS .
• UNIX:
Depending on your release of SAS 9.4, add your CA root certificate and intermediate certificates either to the trusted CA bundle or to the SAS Private
JRE on each of your middle-tier machines.
For more information, see “Add Your Certificates to the Trusted CA Bundle” in
Encryption in SAS
or “Add Your Certificates to the SAS Private JRE” in
Encryption in SAS
.
15. On each of your middle-tier machines, start SAS Deployment Wizard by navigating to
SAS-installation-directory/SASDeploymentManager/9.4
, and launch
setup.exe
(Windows) or
setup.sh
(UNIX). On Windows, you can use the shortcut on the Start menu.
16. On the Select Deployment Type page, select Perform a Planned Deployment and
Configure SAS Software.
Note: Make sure that Install SAS Software is deselected.
17. When prompted for the configuration prompting level, select Custom.
18. On the SAS Web Server Configuration page, select HTTP Protocol using Secure
Sockets, and click Next.
19. The deployment wizard prompts you for the following information. Enter the values
that you recorded earlier in Table 14.3
, and click Next:
• In the X.509 certificate field, enter the path to the X.509 certificate with the DNS name of the current machine as the common name.
How to Enable IOM Communication over HTTPS Proxy
219
• In the RSA private key field, enter the path to the RSA private key that is not protected by a passphrase.
Note: SAS Web Server does not use a passphrase so that it can start without
having to prompt the administrator for a passphrase.
20. When prompted by the deployment wizard, select Generate a default JKS format
keystore from SAS Environment Manager, and click Next.
T I P
We recommend that instead of providing your own certificates to SAS
Environment Manager during installation, that you accept the default and let SAS
Deployment Wizard install auto-generated self-signed certificates for you. After the deployment wizard is finished, you can go back and replace the default SAS
Environment Manager auto-generated JKS keystore with certificates that you provide. Choosing Generate a default JKS format keystore from SAS
Environment Manager saves you several manual configuration steps.
Using a Combination of Auto-Generated and Your Own Certificates
For SAS Deployment Agent and its remote client and for the SAS Environment Manager server and its agents, SAS Deployment Wizard provides auto-generated, self-signed certificates by default. Because these certificates have intermachine connections, are behind your firewall, and do not involve users, there is less of a reason to implement site-signed or third-party-signed certificates. For the other middle-tier components—the
SAS Web Server and the SAS Web Application Server—you provide your own certificates.
The steps for using a combination of certificates are almost identical to the steps performed when you provide your own certificates, except that you replace the auto-
generated certificates with your own certificates. For more information, see “Providing
.
How to Enable IOM Communication over HTTPS
Proxy
Introduction
Starting with the third maintenance release of SAS 9.4, you can configure Integrated
Object Model (IOM) communication over HTTPS. You can update a SAS Web Server to act as a HTTPS proxy, allowing secured IOM communications between applications and back end IOM Servers, such as SAS Metadata Server or SAS Workspace Server.
Prior to configuring the proxy, SAS Web Server must be configured for TLS. If you did not choose to configure with secure sockets during the initial installation and configuration with SAS Deployment Wizard, you can manually configure SAS Web
Server to use HTTPS. For more information, see “Configuring SAS Web Server
Manually for HTTPS” in SAS Intelligence Platform: Middle-Tier Administration Guide .
Configuring the HTTPS Proxy
Starting with the fourth maintenance of SAS 9.4, the HTTP component that is shipped by SAS can be configured as an HTTPS proxy. Prior to the fourth maintenance of SAS
220
Chapter 14 • Encryption Tasks
9.4, customers who want to configure an HTTPS proxy need to perform the following steps to update specific httpd components and configuration files:
1. An updated version of mod_proxy_connect is required for SAS Web Server to act as a HTTPS proxy. This update is available from SAS Technical Support. After you receive the update, complete the following: a. Backup the mod_proxy_connect module.
b. Copy the mod_proxy_connect module from SAS Technical Support to the
SASHome\SASWebServer\9.4\httpd-2.2.31.1-64/modules
directory, replacing the existing module if it is in the directory.
Note: Your version of httpd may differ, based on your release of SAS 9.4.
c. Edit the
SAS-configuration-directory\Levn\Web\WebServer\conf
\httpd.conf
file to load the new mod_proxy_connect, by specifying the following:
LoadModule proxy_connect_module
“path/mod_proxy_connect.so”
Note: If SAS Web Server was already loading the module, the statement may
already be defined in the httpd.conf file.
2. Edit the
SAS-configuration-directory\Levn\Web\WebServer\conf
\sas.conf
file and specify which ports the proxy server should allow to connect to the back end by including the following information:
ProxyRequests On
ProxyVia On
# Below should include ports to be proxied, for example, 8561 for the metadata server.
AllowCONNECT 8561 8591
<Proxy *>
Order allow,deny
Allow from all
</Proxy>
Note: The comment must be on one line. It is shown on more than one line in the
preceding code sample for display purposes only.
3. Restart the SAS Web Server by running the following command:
SAS-configuration-directory\Levn\Web\WebServer\bin\httpdctl restart
4. Set the proxy list for your client. Proceed to either
Via Proxy While Using an Environment Variable” on page 221 or “Registering
Proxy Information in SAS Metadata Server For Other Servers” on page 221
.
CAUTION:
Do not use both methods to set the proxy list for your client.
If the environment variable is used and the proxy information stored in metadata, some clients may see this as multiple proxies. It is possible to go through multiple proxies to reach the back end IOM server, therefore configuring both methods may cause the client to attempt to go through the same proxy twice. If the proxy allows connections to itself, this will result in an inefficient trip back through the same proxy. However, if the proxy does not allow connections to itself, this could result in a connection error.
How to Enable IOM Communication over HTTPS Proxy
221
Connecting to an IOM Server Via Proxy While Using an Environment
Variable
Now that the proxy is configured, clients must be directed to use the proxy to communicate with the IOM Server. In the following procedure, we use SAS Metadata
Server and SAS Management Console. An environment variable can be used to force the client application to go through a proxy when making the outgoing connection.
1. The client application’s IOM server connection information should not be changed.
It should appear to continue to connect to the same host and port that the server itself is running on. If the SAS Metadata server is running on
myMetadataServer.mydomain.com on port 8561, then the connection information in
SAS Management Console should be the same:
• Machine: myMetadataServer.mydomain.com
• Port: 8561
2. Configure SAS Management Console to connect through the proxy https://
myproxy.mydomain.com by setting the SAS_IOM_PROXYLIST environment
variable.
Set the variable before SAS Management Console is started. The value of this variable is a semicolon separated list of one or more proxies that the client application must go through before connecting to SAS Metadata Server. The variable is similar to the following:
SAS_IOM_PROXYLIST = https://myproxy.mydomain.com
After this variable is set, all connections to IOM servers from this host will go through the proxy at https://myproxy.mydomain.com. A tunnel through this proxy will be created to communicate with the back end server. In the case of our example, that would be SAS
Metadata Server.
Registering Proxy Information in SAS Metadata Server For Other
Servers
It is also possible to store proxy information in SAS Metadata Server. Applications that find server definitions in SAS Metadata Server, such as the SAS Workspace Server, will use the proxy information stored with the server definition to make a connection through the proxy to the server endpoint.
In SAS Metadata Server, individual machines are configured to use proxies using a proxy list object. Machines are registered in the proxy list object and are mapped to a list of proxies. Any client that is making a connection to the registered machines of the proxy list should find the list of proxies and make connections through these proxies to the machine registered.
To configure the proxy list in SAS Metadata Server, follow these steps.
1. Log on to SAS Management Console as an administrator.
2. On the Plug-ins tab, navigate to Environment Management
ð
Server Manager.
3. Right-click Server Manager, and select New Server.
4. In the New Server Wizard dialog, navigate to and select the Proxy List entry, and then click Next.
222
Chapter 14 • Encryption Tasks
5. Provide a name for the proxy list object and click Next. The name that you specify is for reference only.
6. To add a proxy to the list, click New under the Proxies section. Enter the proxy URI, and then click OK.
7. Repeat step 6 to add all of the proxies. The proxy list is ordered, and applications will enforce the order of the proxies as they appear in this list.
8. After the proxies are added, select the machines to register with the Proxy List. In the
Proxied Machines section, move one or more machines from the Available list to the Selected list. Click Next to continue.
9. A summary of the details about the Proxy List object that will be created is displayed. Click Finish.
After the proxy list is defined in metadata, any connection that is going to an IOM
Server on the defined server (for example, myServer.mydomain.com) will go through the https://myproxy.mydomain.com proxy first.
223
Part 5
Appendix
Appendix 1
Checklists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Appendix 2
User Import Macros . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Appendix 3
Security Report Macros . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Appendix 4
Choices in Workspace Server Pooling . . . . . . . . . . . . . . . . . . . . . . . . . . .
224
225
Appendix 1
Checklists
Checklist for a More Secure Deployment
Introduction
You can use this appendix to verify that security issues are being addressed as appropriate for your environment and goals. Not all measures are relevant in all deployments. It is a good practice to review your security configuration on a periodic basis.
General
• Make sure that the configuration directories are protected by appropriate operating system permissions. See “Recommended Operating System Protections for Windows
Machines” in SAS Intelligence Platform: Installation and Configuration Guide .
• If your deployment includes sensitive data, review host access to SAS data. See
Chapter 8, “Access to SAS Data,” on page 89
.
• On Windows, minimize availability of the privilege Log on as batch job. See
“Windows Privileges” on page 19 .
• Make sure that you are using an encryption model that is appropriate for your environment.See
Chapter 13, “Encryption Model,” on page 183
.
•
Consider limiting the scope of trusted peer connections. See “Trusted Peer
226
Appendix 1 • Checklists
• Consider tracking changes to server logging levels. See “Audit Server Logging Level
Changes That Originate in Client Applications” in SAS Intelligence Platform: System
Administration Guide
.
Web-Layer
For a full discussion of middle-tier security, see
SAS Intelligence Platform: Middle-Tier
Administration Guide
and
SAS Intelligence Platform: Web Application Administration
Guide
.
The following list highlights selected measures:
• Consider disabling concurrent logon sessions to the SAS middle tier. See “Disabling
Concurrent Sign-in Sessions” in SAS Intelligence Platform: Middle-Tier
Administration Guide
.
• Consider reducing the HTTP Session time-out interval for the SAS web applications.
See “Configuring the HTTP Session Time-out Interval” in SAS Intelligence
Platform: Middle-Tier Administration Guide
.
• If your deployment includes sensitive data, make sure that the SAS Web Report
Studio query cache library (wrstemp) and distribution library (wrsdist) are protected by appropriate operating system permissions. See “Manage Host Access to the
Query Cache Directory” in SAS Intelligence Platform: Web Application
Administration Guide
and “Verifying Permissions for the Distribution Library ” in
SAS Intelligence Platform: Web Application Administration Guide
.
• If your deployment includes the SAS Anonymous Web user, determine whether it is necessary and appropriate to keep that identity in place. See
.
Metadata-Layer
• Limit the WriteMetadata permission on servers. See
“Protect Server Definitions” on page 81
.
• Limit the WriteMetadata permission on ACTs. In general, only the SAS
Administrators group has a grant of the WriteMetadata permission on the
Authorization tab of an ACT.
• Limit the WriteMetadata permission on custom folders. To reduce the chance of inadvertent or deliberate changes to a custom folder, grant WriteMemberMetadata
(instead of WriteMetadata) to users who should contribute only content. See
“WriteMetadata and WriteMemberMetadata” on page 31 .
• Review the WriteMetadata permission on OLAP schemas and libraries. To prevent someone from adding cubes to an OLAP schema or tables to a library, set denials of the WriteMetadata permission on the schema or library. Remember to preserve access for administrators as appropriate.
• Review the permission pattern of the predefined ACTs. See
.
• Review who has privileged user status from metadata memberships. See
“Members of Selected Groups and Roles” on page 227
.
•
Consider tracking changes to metadata layer permissions and ACTs. See “Auditing” on page 3
.
Members of Selected Groups and Roles
227
• Consider limiting the locations from which SAS Web Report Studio uses information maps. See “Limit the Availability of Relational Information Maps That Implement
Row-Level Security” in SAS Intelligence Platform: Web Application Administration
Guide
.
Password Protection
• Upgrade encryption strength for passwords in configuration files. See
Maximize Encryption of Stored Passwords” on page 198
.
•
Upgrade encryption strength for outbound passwords in transit. See “How to Change
Encoding for Data in Motion for Outbound Passwords” on page 205 .
• Consider preventing users from saving their passwords in their desktop connection profiles. See
“Client-Side Storage of Passwords” on page 16 .
•
Consider strengthening password policies for internal accounts. See “How to Change
Internal Account Policies” on page 177
.
Members of Selected Groups and Roles
The following table documents the standard members of selected metadata-layer groups and roles.
Table A1.1 Members of Selected Groups and Roles
Group or Role Usual Members
SAS General Servers SAS Trusted User (no others)
SAS System Services
SAS Administrators
Metadata Server: Unrestricted
Metadata Server: Trusted Service Identities
Metadata Server: User Administration
Visual Analytics: Analysis
Visual Analytics: Basic
Visual Analytics: Report Viewing
SAS Trusted User (no others)
SAS Administrator and other administrators
SAS Administrator and possibly other administrators
SAS Trusted User (no others)
SAS Administrators
Visual Analytics Users
SASUSERS
Visual Analytics Users
CAUTION:
228
Appendix 1 • Checklists
If the SAS Trusted User is unrestricted, then problems (such as the inability to launch certain servers) occur. Do not make the SAS Trusted User unrestricted.
Permission Patterns in Predefined ACTs
The following tables document the permission patterns of selected predefined ACTs within the metadata authorization layer. Each row documents the entire pattern for a particular ACT.
The predefined ACTs in the following table are used in the initial configuration and should not be modified.
Table A1.2 Permission Patterns for Selected Predefined ACTs
Denials Grants
ACT Name
Default ACT
Private User Folder ACT
SAS Administrators Settings
*
SAS System Services (group)
PUBLIC
All permissions
-
RM, WM, CM
SAS Administrators
RM, WM, CM, A
RM, WM, CM
RM, WM, CM, A
SASUSERS
-
-
RM, WM. CM
For permission definitions and abbreviations, see “Use and Enforcement of Each
.
SSS
*
RM, WM
RM
RM
229
Appendix 2
User Import Macros
Overview of User Bulk Load and Synchronization . . . . . . . . . . . . . . . . . . . . . . . . .
About the Sample Code for UNIX /etc/passwd . . . . . . . . . . . . . . . . . . . . . . . . . . . .
About the Sample Code for Active Directory . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Location of the User Bulk Load and Synchronization Macros . . . . . . . . . . . . . . .
Overview of User Bulk Load and Synchronization
In order to make access distinctions and track user activity, the metadata server has to know who is making each request. To enable the metadata server to make this determination, each user's metadata definition includes that user's account ID from your authentication provider. The metadata server maintains its own copy of each ID. The metadata server does not maintain copies of external passwords for identification purposes.
Note: For a few administrators, a SAS internal account can be used instead.
This chapter helps you use autocall macros and sample code that SAS provides to create your own programs that bulk load and manage user information. The chapter emphasizes
230
Appendix 2 • User Import Macros
coordination with an Active Directory server or UNIX
/etc/passwd
files but also provides information to help you extrapolate to other providers.
The following figures introduce the batch processes for identity information. In the figures, the
MDU****
items are macros and the libraries contain SAS data sets.
The initial import extracts identity information from your authentication provider and loads that information into the metadata.
Figure A2.1 Initial Import (Bulk Load)
External Enterprise
Identity Source
SAS Metadata
Repository
source-specific extraction cod e
for example: importad.sas
or importpw.sas
Enterprise
Extract Library
(raw data)
source-specific no rmalizing cod e
%MDUIMPLB
(load)
%MDUIMPC
(creates table definitions for canonical tables)
Enterprise
Extract Library
The synchronization performs two extractions (one from your authentication provider and another from the SAS metadata) and then loads validated updates into the metadata.
The numbers in the following figure correspond to these activities:
1. Extract information from your authentication provider.
2. Extract information from the SAS metadata.
3. Compare the two sets of tables and identify updates that need to be made to the metadata (excluding any exceptions metadata that you want to preserve).
4. Validate the changes to make sure that they will not violate the metadata server's integrity constraints.
5. Load the updates into the metadata.
Overview of User Bulk Load and Synchronization
231
Note: Notice that the first part of the import process (the extraction from your
authentication provider) is the same as the first part of the synchronization process.
You will reuse your import extraction code in your synchronization program.
Figure A2.2 Periodic Synchronization
External Enterprise
Identity Source source-specific extraction code
SAS Metadata
Repository
1
Enterprise
Extract Library
(raw data) source-specific normalizing code
%MDUIMPC
(creates table definitions for canonical tables)
Master
Library
Exceptions List
3
%MDUCMP
(comparison)
2
%MDUEXTR
(extract)
Target
Library
Changes
Library
4
%MDUCHGV
(integrity check)
5
%MDUCHGLB
(load changes)
Errors List
232
Appendix 2 • User Import Macros
The following two topics document the format of the data sets and explain how corresponding identity entries are mapped between your authentication provider and the
SAS metadata.
Canonical Tables
The following figure illustrates the names and structures for the tables in which identity information is stored during batch processing:
Figure A2.3 Table Structures and Relationships
email keyid (foreign key) emailAddr emailType grpmems grpkeyid (foreign key) memkeyid (foreign key) location keyid (foreign key) locationName locationType address city postalcode area country person keyid (primary key) name displayName description title idgrps keyid (primary key) name displayName description grpType phone keyid (foreign key) phoneNumber phoneType logins keyid (foreign key) userid password authDomKeyID (foreign key) authdomain keyid (primary key) authDomName
OutboundOnly
TrustedOnly
Here are some key points about the tables:
• You do not have to use all of the tables. For example, if you are not going to store email addresses, you do not need an email table.
• For each table that you do use, all columns must be present. However, you do not have to include data in every column. In the figure, the required columns for the primary objects are indicated with a star.
Note: Each user should have a login that includes a qualified user ID.
• The keyids in the person table (users), idgrps table (groups and roles), and authdomain table (authentication domains) tie each of those primary objects to its related information. In the metadata, the keyid value is stored as an external identity.
For each keyid column, use a fixed, enterprise-wide identifier such as these:
• In the person table, consider using employee identification numbers.
• In the idgrps table, consider using group names (or LDAP Distinguished Names).
• In the authdomain table, consider using authentication domain names.
• All of the relationships between a primary object and its related data are zero-ormore relationships. For example, you can store no phone numbers, one phone number, or multiple phone numbers for each user.
Canonical Tables
233
• We recommend that you avoid using spaces or special characters in the name of a user, group, or role. Not all components support spaces and special characters in identity names.
The following figure depicts data for a user named Tara O'Toole. The ovals indicate personal data for Tara. The check marks indicate data that is indirectly related to Tara
(through her group memberships).
Figure A2.4 Example: Partial Tables Showing Selected User Data
person ke yid
0018
1390
51 07 nam e jhein hhigh totoo di splayNa m e
John Heinsler
Harry Highpoint
Ta ra O’Too le logins keyi d
0018
1390
5107
5107
ETL email keyi d
0018
0018
1390
5107
5107 grpmems grpkey id
Developers
Developers
ETL
ReportConsumers
ReportConsumers memk eyid
Applications
ETL
510 7
Developers
Executives idgrps ke yid
Applications
Developers
ETL
Executives
ReportConsumers name
Applications
Developers
ETL
Executives
ReportConsumers grpT ype em ailA dd r [email protected]
to too @ o rion spo rts .com
otoole@hotm ail.co m emai lT yp e work home work wo rk ho m e useri d
OrionSP\jhein
OrionSP\hhigh
O rionSP\totoo
Tara diadmin authdomKeyid
DefaultAuth
DefaultAuth
DefaultAuth
O racle A uth
DB2Auth authdomain key id
DB2Auth
DefaultAuth
OracleAuth authD o mNam e
DB2Auth
DefaultAuth
OracleAuth
In the figure, notice these things:
• In the person table, the employee number is used as the keyid. The name column uses user IDs (that column also requires unique values). The displayName column contains a common name for each user.
• In the grpmems table (group memberships), Tara is a direct member of the ETL group. Tara is an indirect member of the Developers group, because that group has the ETL group as one of its members. Tara also has a third-level membership in the
ReportConsumers group.
• In the idgrps table, the group name serves as the keyid. This choice is appropriate because the metadata server enforces uniqueness across group and role names.
• In the idgrps table, the grpType column is empty. This indicates that the entries in this table are all groups. To create a role in the metadata, provide a grpType value of
ROLE
.
• In the email table, Tara has two e-mail addresses and each one has a different type.
• In the logins table, Tara has personal logins in two different authentication domains.
Tara can also use the ETL group's login (for DB2Auth), because Tara is a member of the ETL group.
234
Appendix 2 • User Import Macros
• In the authdomain table, the authentication domain name serves as the keyid. This is appropriate because the metadata server enforces uniqueness across authentication domain names.
User Bulk Load
Scope of the Import Process
Who Participates?
In order to participate in the initial import process, an identity must meet both of these criteria:
• The identity must be included in the import tables. If your identity information is distributed across several authentication providers or user registries, extract information from each source and then combine the resulting sets of tables into one set of canonical tables.
To limit the import tables, you can perform these tasks:
• Define a starting point. For example, when you extract identity information from
Active Directory, you specify a Distinguished Name as the starting point. Only identities that exist below that Distinguished Name in the Active Directory hierarchy are extracted.
• Define filters. For example, when you extract identity information from Active
Directory, you can use a filter to extract entries only for people who are members of a particular group.
• Make manual changes to the import tables.
• The identity must not already exist in the SAS environment. You cannot import an identity that has the same name as an identity that already exists in the metadata.
CAUTION:
Do not delete existing SAS identities in order to include them in an initial import. When you delete a SAS identity, you lose that identity's associations
(such as access controls). Creating a new identity with the same name does not restore those associations.
You can incorporate a manually created identity into the synchronization process. To do this, add an external identity on the General tab of that identity's metadata definition.
What Information Is Imported?
The import process can add this information to the metadata:
• user, group, and role definitions with names, display names, descriptions, and membership information
• job titles, contact information, and personal logins for users
Note: In most cases, passwords are not added to the metadata because they typically
cannot be extracted from an authentication provider. If passwords are present in the extracted data, they are loaded into the metadata. It usually is not necessary to include passwords in logins.
Note: Synchronization can process logins for groups. The initial import process does
not support these tasks.
User Bulk Load
235
• authentication domains
Note: Logins in outbound domains will be excluded from all loading and
synchronization updates.
These constraints apply to the initial import:
• When combined with information that already exists in the metadata, the input data must meet uniqueness requirements. For example, you cannot import an identity that has the same name as an identity that already exists in the metadata.
• In order to import a user, group, or role, only a name and one external identity value
(keyid) is required. However, each user should also have at least one login in order to establish an individual SAS identity.
Note: Avoid creating circular or reciprocal group memberships (for example, GroupA is
in GroupB, GroupB is in GroupC, and GroupC is in GroupA). Such relationships introduce unnecessary complexity and can negatively impact performance. When you define nested group memberships (interactively or through importing identity information), the best practice is to use as simple and non-redundant a structure as possible.
How to Import Identities
Note: It is a good practice to run a backup before you perform an import.
To import identity information:
1. Locate the sample code that best fits your external identity source.
•
For import from Active Directory, see “About the Sample Code for Active
• For import from UNIX /etc/passwd files, see
.
• For other formats, the first step is to figure out how to extract the data from your authentication provider. If you have LDAP, you might be able to modify the
Active Directory sample for your purposes. Otherwise, use the %MDUIMPC macro to create empty canonical tables, and then use DATA steps to extract the
information and insert it into those tables. See “Sample Code for User
Synchronization” on page 238 .
2. Decide which attributes you want to add to the metadata. For each attribute, identify a corresponding field in your authentication provider.
3. In the SAS Program Editor, adapt the sample code. The comments in the sample code provide essential details.
4. Submit the code and review the log.
5. In the User Manager plug-in in SAS Management Console, verify that new identities exist. On the General tab of an imported user, group, or role, select External
Identities. You should see an external identity value that matches the identity's keyid in the import tables.
6. Save a copy of your import program for inclusion in your synchronization program.
236
Appendix 2 • User Import Macros
User Synchronization
Scope of the Synchronization Process
Who Participates?
By default, the synchronization process includes all identities that were originally imported into the metadata (because by default only those identities have an external identity in the metadata). You can modify participation in these ways:
• To include an identity that was manually created, add an external identity on the
General tab of that user, group, or role definition.
• To exclude an identity that has an external identity in the metadata, define an exception in the %MDUCMP macro.
What Information Is Updated?
The synchronization process affects the metadata for participating identities as follows:
• Any identities that have been added to your authentication provider are added.
• Any new logins that have been added to your authentication provider are added. If passwords are extracted from your authentication provider, those passwords are included in the new logins.
Note: It is typically not possible to extract passwords from an authentication
provider.
Note: Most logins do not need to include a password.
• Any participating identities that are not found in your authentication provider are deleted.
• By default, only logins that meet all of these criteria are deleted:
• The login belongs to a participating identity.
• The login does not exist in your authentication provider.
• The login is in an authentication domain that exists in your authentication provider.
Note: Logins in authentication domains that do not exist in your authentication
provider are preserved by default.
• If the synchronization includes groups (or roles), memberships for participating SAS identities are updated to match the memberships in your authentication provider. A change in the input grpType value (which determines whether an object is a group or a role) does not cause any update to the metadata.
• Locations, telephone numbers, and e-mail addresses for participating identities are updated. Use exceptions to prevent updates to contact information that is added interactively.
• New authentication domains are added. By default, no authentication domains are removed.
Note: Logins in outbound domains will be excluded from all loading and
synchronization updates.
User Synchronization
237
These constraints apply to updates:
• When combined with information that already exists in the metadata, the change data
must meet all of the metadata server's uniqueness requirements. See “Uniqueness
• In order to add a user, group, or role, only a name and one external identity value
(keyid) is required. However, each user should also have at least one login in order to establish an individual SAS identity.
How to Synchronize Identities
Note: It is a good practice to run a backup before you perform a synchronization (at
least until your program is proven).
To synchronize identity information:
1. If you want to include identities that weren't originally imported, use SAS
Management Console to add a correct external identity value on the General tab of each such identity.
2. If you want to exclude identities or attributes from the update, create an exceptions data set.
CAUTION:
If you used SAS Management Console to make updates to imported identities, those updates are not automatically preserved during batch synchronization.
To preserve such information, define exceptions in the
%MDUCMP macro.
3. In the operating system, set up three directories: an enterprise extract directory, a metadata extract directory, and a change tables directory. For example:
4. In the SAS Program Editor, adapt the sample synchronization code to create your
own program. See “Sample Code for User Synchronization” on page 238
.
5. Submit the code and review the log. To address any errors, make changes in the source tables, the exceptions tables, or the metadata. For details about the errors, examine the errorsds table. After making corrections, run the synchronization program again.
Note: An alternative method for dealing with errors is to re-execute the
%MDUCMP macro with EXCEPTIONS=ERRORSDS. This re-creates the change tables without the offending entries. If an exceptions data set is already being used, you can append the content of the ERRORSDS data set to that data set.
6. In the User Manager plug-in in SAS Management Console, verify that the metadata reflects the changes that you expect to see.
Note: To ensure that current information is displayed, right-click User Manager and
select View
ð
Refresh.
238
Appendix 2 • User Import Macros
Sample Code for User Synchronization
.
/*Specify connection options. Use an unrestricted user ID.*/ options metaserver=machine-name metauser="userID" metapass="password";
/*Specify the directory for the extracted AD data (master tables).
1
*/ libname adir "drive:\path\ADIRextract";
/* Specify the directory for the extracted metadata (target tables).*/ libname meta "drive:\path\METAextract";
/* Specify the directory for the comparison output (change tables).*/ libname updates
"drive:\path\METAupdates";
/* Extract identity information from AD (master).
2
*/
%let _EXTRACTONLY = ;
%include "drive:\path\myimportad.sas";
/* Extract identity information from the metadata (target).*/
%mduextr(libref=meta);
/* Compare AD (master) to metadata (target).
3
*/
%mducmp(master=adir, target=meta, change=updates);
/* Validate the change tables.*/
%mduchgv(change=updates, target=meta, temp=work, errorsds=work.mduchgverrors);
/* Load the changes into the metadata.
4
*/
%mduchglb(change=updates);
The numbers in the preceding code correspond to these points:
1. The AD extract location should match the importlibref location in your import program. For example, if your original importad.sas put the normalized AD data in the work library, you might revise importad.sas by replacing the line:
%let importlibref=work; with these two lines: libname adir "drive:\path\ADIRextract";
%let importlibref=adir;
2. This causes the importad.sas. program to only extract the data (and not perform the load).
3. If you need to exclude items from the comparison, add the EXCEPTIONS=dataset parameter here.
Sample Code for Generic Bulk Load
239
4. If you prefer to not load any changes if %MDUCHGV finds an integrity violation, replace the displayed line with this code:
%macro exec_mduchglb;
%if (&MDUCHGV_ERRORS ^= 0) %then %do;
%put ERROR: Validation errors detected by %nrstr(%mduchgv). Load not attempted.;
%return;
%end;
%mduchglb(change=updates);
%mend;
%exec_mduchglb;
Sample Code for Generic Bulk Load
This code demonstrates one way to import generic identity information. This example is intended to illustrate the structure and relationships of the identity data, rather than to suggest that you enter large quantities of data using this approach. The program uses macro variables to define the tables and uses DATALINES statements to supply input data. The %MDUIMPLB macro creates XML from the tables and submits that XML (in blocks) to the metadata server.
/* Specify connection options. Use an unrestricted user ID. */ options metaserver=machine-name metauser="userID" metapass="password";
/* Initialize the macro variables that create canonical tables. */
%mduimpc();
/* Create the person table. */ data &persontbla ;
%definepersoncols;
infile datalines delimiter=',' missover;
input keyid name description title;
datalines;
P001,Michelle Harrell,Mgr of Operations,Sr. Mgr
P002,Fred Granite,NE Region Acct. Rep.,Account Rep II
P003,Brian Davis,Accounting System Developer,Senior Programmer
;
/* Create the phone table. */ data &phonetbla ;
%definephonecols;
infile datalines delimiter=',' missover;
input keyid phoneNumber phoneType;
datalines;
P001,x1532,Office
P001,(919) 555-1212,Home
P003,x2312,Office
;
/* Create the location table. */
240
Appendix 2 • User Import Macros
data &locationtbla ;
%definelocationcols;
infile datalines delimiter=',' missover;
input keyid locationName locationType address city postalcode area country;
datalines;
P001,My Company,Office,123 Oak Ave,Clayton,20711,CA,USA
P001,Michelle Harrell,Home,105 Seth Ct.,Apex,20765,CA,USA
P002,Fred Granite,Home,2138 Pond St.,Greenlevel,20832,CA,USA
P002,My Company,Office,123 Oak Ave,Clayton,20711,CA,USA
P003,My Company,Office,123 Oak Ave,Clayton,20711,CA,USA
;
/* Create the email table. */ data &emailtbla ;
%defineemailcols;
infile datalines delimiter=',' missover;
input keyid emailAddr emailtype;
datalines;
P001,[email protected],business
P001,[email protected],home
P002,[email protected],business
P003,[email protected],business
;
/* Create the idgrp table. */ data &idgrptbla ;
%defineidgrpcols;
infile datalines delimiter=',' missover;
input keyid name description grpType;
datalines;
G001,Operations Staff,Members of the operations department,
G002,All Groups,Group containing all groups,
G003,Backup Operators, ,
;
/* Create the grpmems table. */ data &idgrpmemstbla;
%defineidgrpmemscols;
infile datalines delimiter=',' missover;
input grpkeyid memkeyid;
datalines;
G001,P001
G002,G001
G002,G003
;
G003,G001
G003,P003
/* Create the authdomain table. */ data &authdomtbla;
%defineauthdomcols;
infile datalines delimiter=',' missover;
input keyid authDomName;
datalines;
A001,DefaultAuth
About the Sample Code for UNIX /etc/passwd
241
A002,UnixAuth
;
/* Create the logins table. */ data &logintbla;
%definelogincols;
infile datalines delimiter=',' missover;
input keyid userid password authdomkeyid;
datalines;
P001,WinNet\Michelle, ,A001
P001,Michelle, ,A002
P002,WinNet\Fred, ,
P003,WinNet\Brian, ,
P003,Brian, ,A002
;
/* Load the information from the work library into the metadata server. */
%mduimplb();
About the Sample Code for UNIX /etc/passwd
Note: This code is in SAS-installation-directory
\SASFoundation\9.4\core
\sample\importpw.sas
(Windows) or SAS-installation-directory
/
SASFoundation/9.4/samples/base/importpw.sas
(UNIX). This topic highlights key points about the code.
This program expects user information in
/etc/passwd
files in this form: user name : password : uid (numeric) : primary group id : <continued> gcos-field: home-directory : login-shell
The
gcos-field
item consists of additional comma-delimited fields in this form:
Person Name, Office, phone extension, misc, employeeid
For example: user name : password : uid (numeric) : primary group id : <continued>
Person Name, Office, phone extension, misc, employeeid:<continued> home-directory : login-shell
Here are details about this code:
• The colons delimit standard fields; the commas delimit items within the
gcosfield
(you can use any delimiter other than a colon).
• If your
/etc/passwd
file has a different format, modify the sample program to either exclude information from the extraction or to extract it from the appropriate fields.
• If the
employeeid
field for an
/etc/passwd
entry is empty, that entry is dropped from the import.
• The values in the
Person Name
field should be unique. The program includes a
DUPLICATEPERSONS macro variable that, when set to RECODE, changes a duplicate value in the
Person Name
field to the user ID value of the PW file entry.
However, if the DUPLICATEPERSONS macro variable is set to any other value, users with duplicate names in the
Person Name
field are deleted.
242
Appendix 2 • User Import Macros
The program expects group information in a
/etc/group
file. The standard format is as follows: groupname : password : gid (numeric) : comma-delimited list of users
Here are details about this file:
• Users are identified by user name (rather than by a numeric user ID).
• A group cannot be a member of another group.
• To exclude a user or group, enter an asterisk (*) in the password field. The import program drops entries that contain an asterisk in the password field.
The following table highlights selected variables:
Table A2.1 Selected Variables in a UNIX File Import
Variable Name
UID
Purpose
Provides an external identity value for each metadata user and group that this program creates.
Notes
Use a field that contains a unique and unchanging value for each entry in an
/etc/passwd
file and an
/etc/group
file. In the sample code, the uid field of the gcos-field is used. If this field is empty for any records, those records are excluded from the import.
MetadataAuthDomain Specify a SAS authentication domain name. The supplied value is typically
DefaultAuth
.
UNIXEMAILDOMAIN Enables construction of an e-mail address for each user.
Specify the UNIX domain for the extracted identities. The supplied value is appended to extracted user IDs to yield e-mail addresses in the form userid@supplied-value. If your e-mail addresses follow a different convention, modify this part of the code.
PWExtIDTAG
Enables all metadata logins that this program creates to be associated with an authentication domain.
Provides a label for all metadata items that this program creates. The label indicates which objects were created by this program.
Specify a descriptive label that will be applied to all imported objects to indicate where they came from. The default value is
Passwd File Import
. Do not quote this value. If you select the External Identities button on an imported identity's
General tab (in SAS Management Console), you will see this label in the Context column of the External Identities dialog box.
About the Sample Code for Active Directory
Note: This code is in SAS-installation-directory
\SASFoundation\9.4\core
\sample\importad.sas
(Windows) or SAS-installation-directory
/
SASFoundation/9.4/samples/base/importad.sas
(UNIX). This topic highlights key points about the code.
Here are some tips for using the program:
About the Sample Code for Active Directory
243
• The code uses the SAS interface to LDAP (the LDAP CALL Routine interface) to extract information from Active Directory.
• The code references standard Active Directory schemas to identify user and group attributes. If your site has extended the standard schema, you might need to make changes in section 3 to reference additional or alternate attributes.
• The code uses filters to segment retrieval. It might be necessary to alter the filters in sections 3 (user extraction) and 4 (group extraction) to better fit the contents of your
Active Directory server. If the number of records in a request exceeds Active
Directory's maximum query limit, only a subset of the requested records is returned.
The Microsoft utility program LDIFDE can be useful in defining appropriate filters.
• If either or both of the data sets that are extracted from Active Directory are empty, execution of the code is canceled and an error message is provided. This reduces the likelihood of inadvertent deletion of metadata identities due to a problem with the
Active Directory extraction during the synchronization process.
Note: Execution is canceled if no users are extracted, and execution is canceled if no
groups are extracted. If you want to extract only users from Active Directory, you must modify the 9.4 sample code. Either disable the error check or remove the entire section of the code that extracts groups.
• The code will not import membership information for a group that has more than
1500 members. (This limitation is version-specific. Check the documentation for your Active Directory server for details). To incorporate an oversize group, use an approach like this:
1. Rewrite the section 4 filters to exclude large groups.
2. Create an additional extraction that uses LDAP range retrieval specifiers to extract the large group membership information in segments. See the
LDAPS_SEARCH CALL routine in the SAS Integration Technologies: Directory
Services Reference.
3. Add that membership information to the main extracted grpmems table.
The following table highlights selected macro variables:
Table A2.2 Selected Macro Variables in an Active Directory Import
Variable Name
keyidvar
Purpose Notes
Provides an external identity value for each metadata user that this program creates.
Specify an LDAP attribute that contains a unique and unchanging value for each user. The sample code uses the employeeID attribute.
*
MetadataAuthDomain Enables all metadata logins that this program creates to be associated with an authentication domain.
Specify a SAS authentication domain name. This value is not related to a Windows domain name. In the standard configuration, the correct value is
DefaultAuth
(or, if you have configured web authentication and are extracting information for users who use only web clients,
web
).
WindowsDomain Enables construction of a qualified user ID in each login that this program creates.
Provide the Windows domain name for the extracted identities.
The supplied value is prepended to each extracted user ID to yield qualified IDs in the form supplied-value\userID.
244
Appendix 2 • User Import Macros
Variable Name
ADExtIDTag
Purpose Notes
Provides a label for all metadata items that this program creates. The label indicates which objects were created by this program.
Specify a descriptive label that will be applied to all imported objects to indicate where they came from. The default value is
Active Directory Import
. Do not quote this value.
If you select the External Identities button on an imported identity's General tab (in SAS Management Console), you will see this label in the Context column of the External Identities dialog box.
*
If this attribute is empty, consider using sAMAccountName or distinguished name.
Location of the User Bulk Load and
Synchronization Macros
The user import and synchronization macros are in SAS-installation-directory
\SASFoundation\9.4\core\sasmacro
(Windows) or SAS-installation-directory
/
SASFoundation/9.4/sasautos
(UNIX). The following topics provide basic reference information that helps you use these macros to perform standard user import and synchronization tasks.
Dictionary
%MDUIMPC
Defines a set of macro variables that create canonical tables and, optionally, creates empty tables for views for CSV files.
Used by:
User import, user synchronization
Syntax
,
);
Optional Arguments
LIBREF
specifies the output location for any canonical tables or views that are created (the default is
Work
). If you choose to not create tables or views, this parameter is not used.
MAKETABLE
controls whether any tables or views are created.
0 specifies that no tables or views are created (this is the default).
1 specifies that empty canonical tables are created.
%MDUIMPLB
245
2 specifies that DATA step views of the CSV files in the fileref directory are created.
INFILEREF
specifies the location of the input CSV source files (used only when
MAKETABLE=2). The CSV input files must have names and contents that match the table names and structures of the canonical tables. On UNIX, the filenames must be lowercase.
FILEHEADERS
indicates whether the CVS input files have a header line (used only when
MAKETABLE=2).
1 indicates that the input files have header information in the first row.
other_value
indicates that the data begins in the first row.
Details
Table A2.3 Tables and Macro Variables Created by %MDUIMPC
Table
person location phone email idgrps
Purpose
Create users
Store addresses
Store phone numbers
Store e-mail addresses
Create groups and roles grpmems Store membership information authdomain Create SAS authentication domains logins Create logins
Name/Keep List Global
Variable
&persontbla;
&locationtbla;
&phonetbla;
&emailtbla;
&idgrptbla;
&idgrpmemstbla;
&authdomtbla;
&logintbla;
Column Attributes
Macro
%definepersoncols
%definelocationcols
%definephonecols
%defineemailcols
%defineidgrpcols
%defineidgrpmemscols
%defineauthdomcols
%definelogincols
%MDUIMPLB
Generates and submits blocks of XML to load identity information.
Used by:
User import
Requirement:
Connection to the metadata server
246
Appendix 2 • User Import Macros
Syntax
,
,
,
<BLKSIZE=number-of-records> );
Optional Arguments
This macro uses PROC METADATA to submit load requests.
LIBREF
specifies the location of the input data. The default is
Work
.
TEMP
specifies the location for temporary data sets during processing. The default is
Work
.
FAILEDOBJS
is a table that contains the entire contents of all blocks that contain any problematic data. The default is
mduimplb_failedobjs
.
This table is populated if integrity and completeness requirements are not met or if a metadata server request failure occurs during processing.
EXTIDTAG
indicates the origin of identity information. The default for this string is
IdentityImport
.
BLKSIZE
specifies the number of identities in a block. Each block of XML is generated and submitted to the metadata server independently. The default is 100.
Details
Table A2.4 FAILEDOBJS
Column
tablename chgtype objtype
Name
Objid
KeyId
UserId
AuthDomKeyId
MemObjType
*
MemName
*
MemObjID
*
Contents
Name of an intermediate input table
A (add), U (update), or D (delete)
Internal metadata object type
Name of a user, group, or role
Internal object ID
Key ID for an item
User ID value in a login
Key ID of an authentication domain
Internal metadata object type for a member
Person indicates a user, IdentityGroup indicates a group
Internal metadata object ID
%MDUCMP
247
Column
MemKeyId
*
Contents
Key ID of a member
*
This column is populated only if there is an error in membership information.
Note: Logins in outbound domains are excluded from all loading and synchronization
updates.
%MDUEXTR
Extracts identity information from the metadata.
Used by:
User synchronization
Requirement:
Connection to the metadata server
Syntax
%MDUEXTR (
Required Argument
This macro uses PROC METADATA to submit extraction requests.
LIBREF
specifies the location to which the metadata is extracted.
%MDUCMP
Generates data sets that contain the changes that must be made to the metadata.
Used by:
User synchronization
Syntax
,
,
<EXCEPTIONS=<libref.>data-set>
,
,
);
Required Arguments
MASTER
specifies the location of the master tables (use the libref that you specify in
%MDUIMPC).
TARGET
specifies the location of the target tables that contain information extracted from the metadata (use the libref that you specify in %MDUEXTR).
CHANGE
specifies the location for the change tables. These tables are created (xxx is the base name of each canonical table):
248
Appendix 2 • User Import Macros
xxx_add
xxx_update
contains users, groups, and roles to be added to the target tables to make them look like the master tables.
contains users, groups, and roles to be modified in the target tables to make them look like the master tables.
xxx_delete
contains users, groups, and roles to be deleted from the target tables to make them look like the master tables.
summarizes changes to users (Person objects).
person_summary idgrps_summary summarizes changes to groups and roles
(IdentityGroup objects).
authdomain_summary summarizes changes to SAS authentication domains
(AuthenticationDomain objects).
Optional Arguments
EXCEPTIONS
specifies a data set that contains exception values.
EXTERNONLY
defines the scope of the comparison. Unless the master data set has an ObjectId column, this option has no effect. A typical master data set does not include an
ObjectId column. A master data set that is extracted from the SAS Metadata
Repository (rather than from your authentication provider) does include an ObjectId column. Extraction of a master data set from the metadata repository happens in the identity synchronization processes for some solutions.
1 specifies that only identities that have an external identity value are included in the comparison. This is the default value.
0 specifies that all identities are included in the comparison.
If EXTERNONLY=1 but AUTHDOMCOMPARE=NAME, all authentication domains are compared. In other words, for authentication domains
AUTHDOMCOMPARE=NAME overrides EXTERNONLY=1.
AUTHDOMCOMPARE
defines how authentication domains are compared.
NAME compares all authentication domains by name. Prevents deletion and renaming of all authentication domains. Prevents deletion of logins in authentication domains that do not exist in the master data set. This is the default.
KEYID compares by keyid. Can cause deletion of authentication domains that were originally imported but are not present in the master data set.
Can cause renaming of authentication domains that were originally imported but have a different name in the master data set. Does not prevent deletion of logins in authentication domains that do not exist in the master data set. If you specify
AUTHDOMCOMPARE=KEYID, authentication domains and logins that are interactively created might be deleted (for a standard synchronization, do not use AUTHDOMCOMPARE=KEYID).
Details
The exceptions data has these columns:
%MDUCHGV
249
tablename
specifies the name of the canonical table to which the exception applies. Valid values are
person
,
logins
,
,
phone
,
location
,
idgrps
,
grpmems
, and
authdomain
.
filter
specifies a SAS WHERE clause expression (without the WHERE) to apply against the corresponding table. The WHERE clause consists of a canonical table column name and an exception value.
For example, consider this exceptions data set: phone PhoneType="manual Phone" email EmailType="manual Email" logins authDomKeyId="A002" logins userid="testid%"
Each line protects a set of objects in a particular target table, ensuring that those metadata objects are preserved.
• The first entry excludes objects in the target phone table that have a PhoneType of
manual Phone
. In SAS Management Console, the PhoneType is displayed in the
Type field in the Phone Properties dialog box.
• The second entry excludes objects in the target email table that have the EmailType value
manual Email
. In SAS Management Console, the EmailType is displayed in the Type field in the Email Properties dialog box.
• The last entry excludes any objects in the target logins table that have a userid value that begins with
testid
.
Note: Logins that are in authentication domains that do not exist in the master tables are
preserved by default. It is not necessary to define exceptions for such logins.
Note: Logins in outbound domains are excluded from all loading and synchronization
updates.
%MDUCHGV
Checks the change tables against the target tables to ensure that the updates do not introduce any integrity problems.
Used by:
User synchronization
Syntax
,
<CHANGE=libref> , <TEMP=libref> ,
Required Argument
TARGET
specifies the location of the target canonical tables. This is typically the same libref that you specify in the %MDUEXTR macro.
250
Appendix 2 • User Import Macros
Optional Arguments
CHANGE
specifies the location of the change tables (that were populated by the %MDUCMP macro).
TEMP
specifies the location for temporary tables (the default is
Work
).
ERRORSDS
identifies a data set that contains information about any integrity problems. This data set has these columns:
errcode errmsg tablename filter
Keyid
Name userid
specifies a numeric code for a particular error.
describes a particular error.
specifies the name of the canonical table from which a particular error item should be excepted if the ERRORSDS data set is fed into the %MDUCMP macro as the exception data set.
specifies a SAS WHERE clause that is used to remove all objects related to a particular error from the change tables.
specifies the keyid value of the offending object.
specifies the Name value of the object, if the offending object is a Person or IdentityGroup.
specifies the userid value of the object, if the offending object is a Login.
authDomKeyId
specifies the keyid value of an authentication domain when duplicate userid values are found in an authentication domain.
If any errors are detected, this macro sets the DUCHGV_ERRORS column to a nonzero value and creates the ERRORSDS data set, which lists the errors that were found.
Note: Logins in outbound domains are excluded from all loading and
synchronization updates.
%MDUCHGLB
Generates XML for loading identity updates.
Used by:
User synchronization
Requirement:
Connection to the metadata server
Syntax
,
<TEMP=libref> , <FAILEDOBJS=table-name> ,
<EXTIDTAG=context-value> , < <BLKSIZE=number-of-records>
);
%MDUCHGLB
251
Required Argument
CHANGE
specifies the location of the change tables.
Optional Arguments
TEMP
specifies the location for temporary tables. The default is
Work
.
FAILEDOBJS
is a table that contains the entire contents of any blocks that were not successfully loaded into the metadata. An error within a block prevents loading of all of the data in that block. The default name for this table is
mduchglb_failedobjs
.
This table is populated if integrity or completeness requirements are not met or if a metadata server request failure occurs during processing. To address errors, see
“How to Synchronize Identities” on page 237
.
EXTIDTAG
Indicates the origin of identity information. The default for this string is
IdentityImport
.
BLKSIZE
specifies the number of added or updated objects in a block. Each block of XML is generated and submitted to the metadata server independently. The default is 100.
For Delete operations, BLKSIZE has no effect.
Note: Logins in outbound domains are excluded from all loading and
synchronization updates.
252
Appendix 2 • User Import Macros
253
Appendix 3
Security Report Macros
Additional Resources for Building Authorization Data Sets . . . . . . . . . . . . . . . . .
Overview of the Security Report Macros
Authorization reporting creates a snapshot of metadata layer access control settings and uses that snapshot in these ways:
• as a data source for reports about current settings
• as a data point for comparing settings across time
The first task in security reporting is to extract, filter, and format authorization data for a specified set of identities, permissions, and objects. SAS provides macros that help you perform this task. For example, the following code uses the main security report macro,
%MDSECDS, to generate authorization data sets for a specified folder and its contents:
/* connect to the metadata server */ options
metaserver=machine-name
metauser="sasadm@saspw"
metapass="{SAS002}3CD4EA1E35CA49324AOC4D63";
/* run the main report macro against a target folder */
%mdsecds(folder="\demo"); run;
The second task in security reporting is to build reports that run against the authorization data sets. For example, the following code prints some of the data from the preceding example:
/* print the WriteMetadata information */ proc print data=work.mdsecds_join noobs; var objname publictype identityname WriteMetadata; run;
The output looks like this:
254
Appendix 3 • Security Report Macros
Because the output is SAS data sets, you can create sophisticated reports by using SAS reporting techniques such as these:
• create an information map that uses an authorization data set as its data source and then use SAS Web Report Studio to create reports based on that information map.
• write SAS ODS (output display system) code that builds an HTML report against the authorization data sets. For an example, see SAS-installation-directory
\SASFoundation\9.4\core\sample\secrpt.sas
(Windows) or SAS-
installation-directory
/SASFoundation/9.4/samples/base/secrpt.sas
(UNIX).
For details, options, and examples, see
Authorization Data Sets
For reference, this topic documents the structure of the generated tables.
Note: The output columns for Permissions will have underscores instead of spaces.
Table A3.1 work.mdsecds_join (the Primary Table for Reporting; Combines objs, permsw, pconds)
Column Name Column Contents
IdentityDispName
The display name of this user or group (for example,
SAS Demo User
). If there is no display name, this column contains the same value as the IdentityName column.
Authorization Data Sets
255
Column Name
IdentityName
IdentityType
Location
MetadataCreated
MetadataType
MetadataUpdated
ObjId
ObjName
ObjUri
ParentObjId
Permissions
(permissions)
PublicType
Column Contents
The name of this user or group (for example,
sasdemo
).
Person
,
IdentityGroup
, or
Role
.
A container path for this object (for example, \Shared Data\Sales). The path usually consists of folder names but can also include names of other containers.
The date on which this object was created.
The internal metadata type for this object.
The most recent date that this object was updated.
The metadata ID for this object (for example,
A5HDAJSI.B900066J
).
The name of this object (for example,
Shoe Sales by Region)
.
The uniform resource identifier for this object (for example,
omsobj:PhysicalTable/
A5XT9KUX.B80000001
).
The metadata ID for this object's immediate parent.
A list of the permissions to inspect for this object. If the list is blank, then all permissions that are applicable to this type of object are inspected.
A separate column for each permission. Each cell in these columns contains a value that indicates whether this permission is effectively granted or denied for this identity. The type of setting
(explicit, ACT, or indirect) is also indicated (for example,
Granted by ACT
or
Denied
Explicitly
or
Granted Indirectly
).
The public metadata type for this object.
Table A3.2 work.mdsecds_objs (Contains Folder and Member Information)
Column Name
Desc
Location
MetadataCreated
MetadataType
MetadataUpdated
ObjId
Column Contents
Description information for this object. For most objects, this information comes from the
Description field on the General tab in the object's properties dialog box.
A container path for this object. The path usually consists of folder names but can also include names of other containers.
The date on which this object was created.
The internal metadata type for this object.
The most recent date that this object was updated.
The metadata ID for this object.
256
Appendix 3 • Security Report Macros
Column Name
ObjName
ObjUri
ParentObjId
Permissions
PublicType
Column Contents
The name of this object.
The uniform resource identifier for this object.
The metadata ID for this object's immediate parent.
A list of the permissions to inspect for this object. If the list is blank, then all permissions that are applicable to this type of object are inspected.
The public metadata type for this object.
Table A3.3 work.mdsecds _pconds (Contains Permission Condition Expressions)
Column Name
Condition
IdentityName
IdentityType
ObjUri
Permission
Column Contents
The expression that limits this grant.
The name of this user or group.
Person
,
IdentityGroup
, or
Role
.
The uniform resource identifier for this object.
The name of this permission (for example,
Read
).
Table A3.4 work.mdsecds_permsw (Contains Permissions Data Transformed to Wide Format)
Column Name
IdentityDispName
IdentityName
IdentityType
ObjName
ObjUri
(permissions)
Column Contents
The display name of this user or group (or the name if there is no display name).
The name of this user or group.
Person
,
IdentityGroup
, or
Role
.
The name of this object.
The uniform resource identifier for this object.
A separate column for each permission. Each cell in these columns contains a value that indicates whether this permission is effectively granted or denied for this identity. The type of setting (explicit, ACT, or indirect) is also indicated.
Additional Resources for Building Authorization Data Sets
257
Table A3.5 work.mdsecds _permsl (Contains Permissions Data in Original Long Format)
Column Name
Authorization
Conditionflg
IdentityDispName
IdentityName
IdentityType
ObjName
ObjUri
Permission
Column Contents
A value that indicates where this permission is granted or denied for this identity. The type of setting (explicit, ACT, or indirect) is also indicated.
A value that indicates whether this permission setting is conditional. The value is either blank
(no condition) or * (condition applies and is written to the pconds table).
The display name of this user or group (or the name if there is no display name).
The name of this user or group.
Person
,
IdentityGroup
, or
Role
.
The name of this object.
The uniform resource identifier for this object.
The name of this permission.
Additional Resources for Building Authorization
Data Sets
This topic introduces the underlying security report macros.
You can choose to directly use the underlying report macros instead of using only
%MDSECDS (which is the standard approach to building authorization data sets).
However, the underlying macros don't offer any unique inclusion or exclusion parameters. The following figure introduces the underlying macros. In the figure, arrow direction indicates input to and output from each underlying macro.
258
Appendix 3 • Security Report Macros
Figure A3.1 Underlying Macros
Metadata Repository
Which folder?
Which objects?
1
%MDSEC G O
work.mdsecd s_ o bjs
Which permissions?
Which identities?
2
%MDSEC
G P
work.mdsecd s_ permsl work. mdsecds _ pco nd s
3
%MDSEC T R
work.mdsecd s_ permsw
4
%MDSEC
VW
work.mdsecd s_ jo in
The numbers in the preceding figure correspond to these activities:
1. %MDSECGO extracts information for a specified set of objects. You specify one folder and indicate whether to include subdirectories. You can also provide a list of object types to include and filter the data set by attribute value.
Note: The same level of control is provided by using %MDSECDS on its own.
2. For every object in a specified data set, %MDSECGP gets effective permission settings for a specified set of identities and permissions.
Note: This is the point at which using the underlying macros creates an opportunity
for you to define a subset of the objs data set.
3. %MDSECTR transforms the extracted data set from a long format (a separate row for each permission) to a wide format (all permissions in the same row).
4. %MDSECVW creates a joined view or data set that can be used for reporting.
The following table introduces a few utility macros that can be useful in security reporting:
%MDSECDS
259
Table A3.6 Utility Macros for Security Reporting
Utility Macro
%MDUTYPED
Description
Extracts information about top-level metadata objects or locates templates for a particular object type.
%MDUGFLDR Returns the object ID for a specified folder.
%DEFINEOBJTAB_SQL Defines the table into which %MDSECGO inserts rows.
For more information, see the macros in SAS-installation-directory
\SASFoundation
\9.4\core\sasmacro
(Windows) or SAS-installation-directory
/
SASFoundation/9.4/sasautos
(UNIX).
Dictionary
%MDSECDS
Generates authorization data sets. This is the top-level macro (it calls the underlying macros and should be used on its own).
Used by:
Security reporting
Type:
Stand-alone
Requirement:
Connection to the metadata server
Syntax
%MDSECDS
,
,
<INCLUDETABLECOMPONENTS=YES | NO>
,
<INCLUDECUBECOMPONENTS=YES | NO> ,
<INCLUDESECUREDTABLES=YES | NO>
,
,
<IDENTITYNAMES="list"> , <IDENTITYTYPES="list">
);
Optional Arguments
OUTDATA
provides a base name for the output. By default, the base name is
work.mdsecds
.
FOLDER
identifies a starting point folder. By default, the starting point is the server root (the
SAS Folders node). If you provide a path (such as
"\Products\SAS
Intelligence Platform\Samples"
), the starting point is the last folder in the
260
Appendix 3 • Security Report Macros
path. To avoid having to type a long pathname, copy the path from a child item's general properties into your code.
INCLUDESUBFOLDERS
controls whether the entire subtree is included. By default, all objects in the entire subtree are included. If you specify
NO
, only the immediate contents of the starting point folder are included.
INCLUDETABLECOMPONENTS
controls whether table columns are included when a table is returned. By default, columns are included.
INCLUDECUBECOMPONENTS
controls whether cube hierarchies, levels, and dimensions are included when a cube is returned. By default, these components are included.
INCLUDESECUREDTABLES
controls whether secured table objects are included when a secured library object is returned. By default, secured table objects are included.
MEMBERTYPES
limits by object type. By default, all public types are included. If you provide a comma-delimited list of types, only those types are included.
You must provide the public type name in its TypeName format. For example, if you access the Advanced tab in the properties dialog box for the Information Map
(relational) object type, you will see that its TypeName is
InformationMap.Relational
.
T I P
In SAS Management Console, all public types are displayed on the Folders tab under System
ð
Types.
Note: If you use this option, examine the output. The log doesn't display errors or
warnings for incorrectly specified types.
MEMBERFILTER
limits by metadata attribute value. By default, no filter is applied. If you provide an expression, only objects that meet the criteria are included. The format for the expression is @attribute-name comparison-operator 'value' (for example,
@ID='A5HDAJSI.B90006Y5'
or
@Name=:'Salary'
).
Comparison operators for character data include
=
(equals),
=:
(begins with),
?
(contains), and
ne
(not equals).
Metadata attributes are associated with an object's metadata type (not public type).
Here are two common attributes:
ID is the object's metadata ID, which is displayed on the object's
Inheritance tab.
Name is the object's name, which is displayed in the Name field on the object's General tab.
To find additional attributes, determine the MetadataType of the object that you are interested in. The type is displayed on the object's Inheritance tab. Then, in the reference documentation for the metadata model, look up that MetadataType to find the names of its attributes.
PERMS
specifies which permissions to include. By default, all permissions that are supported for each object type are included for objects of that type. If you provide a commadelimited list of permissions, only those permissions are included. Even if the
%MDSECDS
261
permissions in your user interface are translated, you must specify the Englishlanguage permission names (for example, ReadMetadata).
IDENTITYNAMES
specifies which identities to include. By default, only the named participants (the identities that are listed in an object's authorization properties) are included. If you provide a comma-delimited list of identity names, only those identities are included.
List identities by their names, not their display names.
If you use this option, you must also use the IDENTITYTYPES option.
IDENTITYTYPES
specifies whether names in the IDENTITYNAMES list correspond to users or to groups. For example, the first name listed in the IDENTITYNAMES parameter must match the first value in the IDENTITYTYPES list. Valid values in this list are
Person
and
IdentityGroup
.
Examples
Example 1: Permissions for All Objects within a Folder
This code extracts information about permissions on the objects in the Sales folder but doesn't include objects in subfolders:
%mdsecds(folder="\Shared Data\Sales", includesubfolders=no);
Example 2: Permissions for Two Object Types within a Folder
Branch
This code extracts information about permissions on tables and schemas in the Sales folder and its subfolders:
%mdsecds(folder="\Shared Data\Sales", membertypes="Library,OLAPSchema");
Example 3: ReadMetadata Permission for Libraries for a Specified
User
This code extracts information that indicates which libraries a particular user (the SAS
Demo User) can see:
%mdsecds(identitynames="sasdemo", identitytypes="Person", membertypes="Library",
perms="ReadMetadata");
Example 4: ReadMetadata Permission for Stored Processes for Two
Specified Users
This code extracts information that indicates which stored processes two users (the SAS
Demo User and Tara O'Toole) can see:
%mdsecds(identitynames="sasdemo,totoo", identitytypes="Person,Person",
membertypes="StoredProcess", perms="ReadMetadata");
Example 5: WriteMetadata Permission for Reports for Specified
Identities
This code extracts information that indicates which reports one user and one group (the
SAS Demo User and PUBLIC) can modify:
%mdsecds(identitynames="sasdemo,PUBLIC", identitytypes="Person,IdentityGroup",
membertypes="Report", perms="WriteMetadata");
262
Appendix 3 • Security Report Macros
Example 6: ReadMetadata Permission for a Subset of Reports
This code extracts information that indicates who can view reports that include the word
"Salary" in their names:
%mdsecds(membertypes="Report", perms="ReadMetadata", memberfilter="@Name ? 'Salary'");
Example 7: Permissions for an Object (Referenced by Object ID)
This code extracts permission settings for an object that is referenced by its object ID:
%mdsecds(memberfilter="@ID='A5HDAJSI.B90006Y5'");
Note: Member filters are not applied to folders, so this example returns all folders (in
addition to the object that has the specified ID).
263
Appendix 4
Choices in Workspace Server
Pooling
Which Eligible Requests Actually Use Pooling? . . . . . . . . . . . . . . . . . . . . . . . . . . .
About Workspace Server Pooling
The primary purpose of pooling is to enhance performance by avoiding the time penalty that is associated with launching all workspace servers on demand. In pooling, a set of workspace servers are made available to process certain types of requests. For example, pooling is used when a relational information map is queried, processed, opened, or used indirectly (through a report).
The initial configuration in a new deployment conforms to the following general recommendations:
• Use server-side pooling. The initial configuration includes a logical pooled workspace server.
• Use client-side pooling only if you have security requirements that are not met by server-side pooling. The initial configuration does not include client-side pooling.
Benefits and Risks of Server-Side Pooling
Server-side pooling offers the following benefits:
• With server side-pooling, the spawner can request user-specific metadata-layer access evaluations, such as whether a user has permission to use a particular server.
This is not possible in client-side pooling because use of the pool administrator prevents the spawner from knowing who the requesting user is.
• With server-side pooling, you can track each request to a specific server instance.
This level of auditing is not possible in client-side pooling because use of the pool administrator prevents the spawner from knowing who the requesting user is.
264
Appendix 4 • Choices in Workspace Server Pooling
• With server-side pooling, testing of an information map in SAS Information Map
Studio more closely approximates results in SAS Web Report Studio, because both requests can use the same server.
• With server-side pooling, allocation of server processes is managed across clients, using the spawner's load-balancing capabilities. Client-side pooling uses round-robin assignments on a per-application basis.
• A side effect of pooling is that the launched workspace servers run under a designated service account. This side-effect is beneficial in avoiding credential gaps that can occur when a desktop application requests a workspace server from a middle-tier service. This benefit is not provided by client-side pooling, because client-side pooling is supported only for web applications.
Server-side pooling introduces the following risks:
• Someone might write a rogue application that bypasses metadata layer access controls. This risk originates from the following difference in server control:
• In client-side pooling, the server is controlled by a designated web application.
Only that application can request a pooled workspace server, because only that application can provide the pool administrator's credentials. No other application can use (or exploit) the server.
• In server-side pooling, there is no pool administrator or alternate form of application-based gatekeeping. Anyone who can use the server legitimately
(from a SAS application) can also exploit that server by writing a rogue application that uses the server to fetch SAS tables. The security issue is that retrieval from such an application bypasses metadata-layer controls that apply when the same data is accessed in a legitimate manner. For example, a request in
SAS Web Report Studio might be filtered by a metadata-layer permission condition that enables the user to see only certain rows in a data set.
Note: This exposure is also applicable to other SAS processing servers that provide
mediated access (standard workspace servers that use SAS token authentication, stored process servers). This exposure occurs in client-side pooling if a user obtains the pool administrator's credentials.
Note: To avoid a similar exposure, nobody is allowed to assign a stored process to a
server-side pooled workspace server.
• Someone might exploit the server-side pooled workspace server from within SAS
Information Map Studio. In general, only information map creators who have host access to the target data use this application. If someone else has SAS Information
Map Studio, that person could bypass metadata-layer controls when querying a relational information map from within that application.
Note: This is never a risk with client-side pooling, because client-side pooling is not
available for desktop applications such as SAS Information Map Studio.
The following table summarizes the trade-offs. For completeness, the table includes a column for a standard workspace server that uses SAS token authentication and a column for a standard workspace server that uses either form of host authentication
(credential-based or Integrated Windows authentication).
Which Requests Are Eligible to Use Pooling?
265
Table A4.1 Summary: Comparison of Workspace Server Configurations
Pooled
Server-
Side
Client-
Side
Not Pooled
SAS
Token Host Advantage
Performance and Compatibility:
Improves performance of web applications.
Compatible with spawner-based load balancing.
Compatible with web authentication.
Accommodates batch generation of scheduled reports.
*
Fully compatible with Integrated Windows authentication.
**
Security:
The SAS identity of the requesting user (or group) is used for metadatalayer evaluations.
Server access security is supported.
The server is controlled by designated client applications.
The user (or group) host identity is passed to the host layer.
The user (or group) host identity is passed to SQL Server.
***
*
**
***
Credentials for the workspace server are available to the batch generation process.
Accommodates requests through SAS Intelligent Query Services for a workspace server (for example, opening an information map after logging on to SAS Enterprise Guide using Integrated Windows authentication).
If the workspace server's host authentication is by IWA and any additional configuration requirements are met.
Which Requests Are Eligible to Use Pooling?
Requests that are handled by a particular query services software component are eligible to use pooling. That software component is primarily used to query, process, open, or otherwise interact with a relational information map.
Note: In SAS Enterprise Guide and the SAS Add-In for Microsoft Office, not all such
requests are eligible. If the libraries that an information map references can be assigned within an existing SAS session, a request to open that information map is not eligible to use pooling. To ensure that such requests are eligible, limit physical
(host operating system) access to the directories that are referenced by that information map's libraries. Deny access to users and grant access to the host identity under which a pooled server runs.
Note: Similar requests that do not involve a relational information map are not eligible,
because those requests are not handled by the query services component. For
266
Appendix 4 • Choices in Workspace Server Pooling
example, requests to open a report that directly contains a stored process or open a report that contains OLAP data are not eligible.
Note: A few specialized low-level tasks (such as dynamically building a list of prompt
values) are eligible, because the query services component is used for those tasks.
Not all requests that can use pooling actually do so. See the following topic.
Which Eligible Requests Actually Use Pooling?
Use of pooling for eligible requests is constrained as follows: client-side pooling can be used for eligible requests in only specially configured web applications. For example, if SAS Web Report Studio's configuration includes a pool administrator, that application uses client-side pooling to process information maps.
server-side pooling can be used for eligible requests in any application. For example, server-side pooling can be used to process information maps from SAS Web Report Studio, SAS
Information Map Studio, SAS Enterprise Guide, and the SAS Add-In for Microsoft
Office. However, eligible requests do not use server-side pooling in the following circumstances:
• If a web application has a configured pool administrator, requests from that application do not use server-side pooling.
• If there is no logical pooled workspace server under a particular application server , requests for resources that are assigned to that application server cannot use server-side pooling.
• If a user does not have the ReadMetadata permission for the logical pooled workspace server , requests made by that user cannot use server-side pooling.
Note: If you chose to limit ReadMetadata permission for the server, be sure to
preserve necessary access for service identities.
Note: If the Use Server Access Security check box on the logical server's
Options tab is cleared, users do not need ReadMetadata permission. This check box should always be selected.
The following figure summarizes the decision sequence that determines what type of server processes an eligible request. Most of the determinative factors can be controlled by an administrator.
Modifying the Initial Pooling Configuration
267
Figure A4.1 Use of Pooling (Applies to Only Eligible Requests)
S tandard S tandard
Did the eligible request originate in a desktop application?
N o
Y es
Is there a logical pooled workspace server?
N o
Y es
N o
Does the originating
Web application have a pool administrator?
Y es
Does the requesting user have RM for the logical pooled server?
Is client-side pooling fully configured?
N o
Y es
N o
Y es
Is the requesting user in a puddle group?
Y es
N o
S tandard
S erve r-side
P o oling
R equest is denied
Cl ient-side
P o o lin g
In the preceding figure, the bold (green) text and arrows mark the decision path in a new deployment with the default configuration. In such a deployment, there are two logical workspace servers within the general-use application server (for example, SASapp):
• a server-side pooled workspace server (the Logical Pooled Workspace Server) that is visible to all registered users.
• a standard workspace server (the Logical Workspace Server) that is not converted to client-side pooling. There is no configured pool administrator, so client-side pooling is not attempted.
Modifying the Initial Pooling Configuration
If you are concerned about the risk that server-side pooling introduces, and that concern outweighs the advantages of server-side pooling, consider these options:
• For a high security implementation of BI row-level permissions, you must use clientside pooling in a separate deployment of SAS Web Report Studio. You can continue to use server-side pooling in the original deployment. See
SAS Guide to BI Row-
Level Permissions
.
• To prevent only web applications from using server-side pooling, configure those applications to use client-side pooling.
Note: In order to get the security benefit, you must set up client-side pooling
correctly. For example, do not use the general SAS Spawned Servers credential
(sassrv) as the puddle login.
• To enable information map creators to run queries under their own host identities
(instead of under the launch credential of the server-side pooled workspace server), hide that server definition from those users. Someone who cannot see the logical pooled workspace server uses the standard workspace server instead.
Note: This is appropriate only if you have an information map creator who should
not be able to access all data that is available to the pooled workspace server.
This does not eliminate the risk of unauthorized use of SAS Information Map
268
Appendix 4 • Choices in Workspace Server Pooling
Studio by someone who legitimately uses the server-side pooled workspace server from other applications.
• To completely eliminate use of server-side pooling, delete the logical pooled workspace server. For performance reasons, we recommend that you also configure client-side pooling for SAS Web Report Studio.
See Also
•
“Hide Server Definitions” on page 84
•
•
“Criteria for a Designated Launch Credential” on page 126
• “Understanding Server Pooling” in SAS Intelligence Platform: Application Server
Administration Guide
Recommended Reading
Here is the recommended reading list for this title:
•
SAS Guide to BI Row-Level Permissions
•
SAS Guide to Metadata-Bound Libraries
•
SAS Intelligence Platform: Application Server Administration Guide
•
SAS Intelligence Platform: Data Administration Guide
•
SAS Intelligence Platform: Desktop Application Administration Guide
•
SAS Intelligence Platform: Middle-Tier Administration Guide
•
SAS Intelligence Platform: Overview
•
SAS Intelligence Platform: System Administration Guide
•
SAS Intelligence Platform: Web Application Administration Guide
•
SAS Management Console: Guide to Users and Permissions
• SAS offers instructor-led training and self-paced e-learning courses to help you administer the SAS Intelligence Platform. For more information about the courses available, see support.sas.com/admintraining.
For a complete list of SAS publications, go to sas.com/store/books . If you have questions about which titles you need, please contact a SAS Representative:
SAS Books
SAS Campus Drive
Cary, NC 27513-2414
Phone: 1-800-727-0025
Fax: 1-919-677-4444
Email: [email protected]
Web address: sas.com/store/books
269
270
Recommended Reading
271
Glossary
access control template (ACT)
a reusable named authorization pattern that you can apply to multiple resources. An access control template consists of a list of users and groups and indicates, for each user or group, whether permissions are granted or denied.
ACT
See
authentication
See
authentication domain
a SAS internal category that pairs logins with the servers for which they are valid.
For example, an Oracle server and the SAS copies of Oracle credentials might all be classified as belonging to an OracleAuth authentication domain.
authentication provider
a software component that is used for identifying and authenticating users. For example, an LDAP server or the host operating system can provide authentication.
authorization
the process of determining the permissions that particular users have for particular resources. Authorization either permits or denies a specific action on a specific resource, based on the user's identity and on group memberships.
capability
an application feature that is under role-based management. Typically, a capability corresponds to a menu item or button. For example, a Report Creation capability might correspond to a New Report menu item in a reporting application. Capabilities are assigned to roles.
client authentication (authentication)
the process of verifying the identity of a person or process for security purposes.
client-side pooling
a configuration in which the client application maintains a collection of reusable workspace server processes. See also
.
272
Glossary
connection profile
a client-side definition of where a metadata server is located. The definition includes a computer name and a port number. In addition, the connection profile can also contain user connection information.
credential management
the reuse of cached credentials or the retrieval of stored credentials from the metadata.
credentials
evidence that is submitted to support a claim of identity (for example, a user ID and password) or privilege (for example, a passphrase or encryption key).
direct LDAP authentication
a configuration in which the metadata server sends credentials to an LDAP provider
(such as Active Directory) for validation, bypassing the host authentication process.
encryption
the conversion of data by the use of algorithms or other means into an unintelligible form in order to secure data (for example, passwords) in transmission and in storage.
external identity
a synchronization key for a user, group, or role. For example, employee IDs are often used as external identities for users. This is an optional attribute that is needed only for identities that you batch update using the user import macros.
host authentication
a process in which a SAS server sends credentials to its host operating system for verification.
Integrated Windows authentication (IWA)
a Microsoft technology that facilitates use of authentication protocols such as
Kerberos. In the SAS implementation, all participating components must be in the same Windows domain or in domains that trust each other.
internal account
a SAS account that you can create as part of a user definition. Internal accounts are intended for metadata administrators and some service identities; these accounts are not intended for regular users.
internal authentication
a process in which the metadata server verifies a SAS internal account. Internal authentication is intended for only metadata administrators and some service identities.
IWA
See
Integrated Windows authentication .
locked-down server
a SAS server that is configured with the LOCKDOWN system option, so that the server can access only designated host resources.
login
a SAS copy of information about an external account. Each login includes a user ID and belongs to one SAS user or group. Most logins do not include a password.
Glossary
273
PAM
See
pluggable authentication module
.
permission condition
a control that defines access to data at a granular level, specifying who can access particular rows within a table or particular members within an OLAP cube. Such controls are typically used to subset data by a user characteristic such as employee
ID or organizational unit.
pluggable authentication module (PAM)
an industry-standard technology that extends UNIX host authentication to recognize additional authentication providers.
puddle
a group of servers that are started and run using the same login credentials. Each
puddle can also allow a group of clients to access the servers. See also client-side pooling .
repository access control template
the access control template (ACT) that controls access to a particular repository and to resources for which access controls are not specified. You can designate one repository ACT for each metadata repository. The repository ACT is also called the default ACT.
restricted identity
a user or group that is subject to capability requirements and permission denials in the metadata environment. Anyone who isn't in the META: Unrestricted Users Role and isn't listed in the adminUsers.txt file with a preceding asterisk is a restricted identity.
role (user role)
a set of capabilities within an application that are targeted to a particular group of users.
SAS authentication
a form of authentication in which the target SAS server is responsible for requesting or performing the authentication check. SAS servers usually meet this responsibility by asking another component (such as the server's host operating system, an LDAP provider, or the SAS Metadata Server) to perform the check. In a few cases (such as
SAS internal authentication to the metadata server), the SAS server performs the check for itself. A configuration in which a SAS server trusts that another component has pre-authenticated users (for example, web authentication) is not part of SAS authentication.
SAS token authentication
a process in which the metadata server generates and verifies SAS identity tokens to provide single sign-on to other SAS servers. Each token is a single-use, proprietary software representation of an identity.
secured data folder
a SAS metadata object that serves as a specialized container for secured library objects, as part of the metadata-bound libraries feature. In each metadata repository, the first secured data folder is Secured Libraries, in the System folder. Additional secured data folders can be added only beneath a Secured Libraries folder.
274
Glossary
secured library object
a SAS metadata object to which a physical SAS library is bound. Metadata-layer permissions on each secured library object manage access to its corresponding physical library. Secured library objects are stored beneath a repository’s Secured
Libraries folder.
secured table object
a SAS metadata object to which a physical SAS data set (a table or view) is bound.
Metadata-layer permissions on each secured table object manage access to its corresponding physical data set. Each secured table object is stored beneath a secured library object.
server-side pooling
a configuration in which a SAS object spawner maintains a collection of reusable workspace server processes that are available for clients. The usage of servers in this pool is governed by the authorization rules that are set on the servers in the SAS metadata.
service identity
an identity or account that exists only for the purpose of supporting certain system activities and does not correspond to a real person. For example, the SAS Trusted
User is a service identity.
single sign-on (SSO)
an authentication model that enables users to access a variety of computing resources without being repeatedly prompted for their user IDs and passwords. For example, single sign-on can enable a user to access SAS servers that run on different platforms without interactively providing the user's ID and password for each platform. Single sign-on can also enable someone who is using one application to launch other applications based on the authentication that was performed when the user initially logged on.
SSO
See
.
trusted user
a privileged service account that can act on behalf of other users on a connection to the metadata server.
unrestricted identity
a user or group that has all capabilities and permissions in the metadata environment due to membership in the META: Unrestricted Users Role (or listing in the adminUsers.txt file with a preceding asterisk).
user context
a set of information about the user who is associated with an active session. The user context contains information such as the user's identity and profile.
user role
See
web authentication
a configuration in which users of web applications and web services are verified at the web perimeter, and the metadata server trusts that verification.
Glossary
275 well-formed user definition
a user definition that includes a login with an appropriate user ID. For a Windows account, the user ID in the login must be qualified (for example, WIN\marcel or [email protected]). The login does not have to include a password. For metadata administrators and some service identities, it is appropriate to use an internal account instead of a login.
276
Glossary
277
Index
Special Characters
%MDSECDS macro
%MDUCHGLB macro
%MDUCHGV macro
%MDUCMP macro
%MDUEXTR macro
%MDUIMPC macro
%MDUIMPLB macro
A
access control templates (ACTs)
See also
consolidation of
permission patterns
predefined
access controls relative priority of
access to data required conditions
access to data and resources opening up
account lockout period
Active Directory
ID format
sample code for user import
administrators adding
SAS Administrators group
align authentication
anonymous access
application features availability of
,
See also
See also
align
coordinating the workspace server
identifying or creating user accounts
token
authentication domains outbound
trusted
unique names and IDs
credential management
direct LDAP authentication
host authentication
Integrated Windows authentication
pluggable authentication modules
SAS internal authentication
SAS token authentication
server type summary
single sign-on summary
trusted peer connections
trusted user connections
authentication model credential gaps
data servers and processing servers
logins
metadata server
mixed providers
PUBLIC and anonymous access
authentication tasks
changing internal account policies
configuring direct LDAP authentication
configuring Integrated Windows authentication
configuring SAS token authentication
configuring web authentication
forcing the Kerberos protocol
storing passwords for third-party server
storing passwords for workspace server
workspace server's Options tab
authorization data sets additional resources for building
generating
authorization model authorization decisions
fine-grained controls for data
granularity and mechanics
identity precedence
inheritance paths
,
278
Index
permissions by item
permissions by task
B
baseline ACTs
examples
,
key points for
batch reporting fine-grained controls and
C
canonical tables
capabilities contributed
explicit
implicit
redistributing
relationship with permissions
CEL parameter
OBJECTSERVERPARMS system option
client-side pooling
contributed capabilities
credential gaps
credential management
credentials launch credentials
D
data access opening up
data retrieval
data servers authentication to
direct access
direct LDAP authentication 136
configuring
dual users
dynamic filters
increasing strength for passwords at rest
encryption for data at rest 184
default settings
184 encryption for data in motion 184
changing settings for SAS servers
default settings
outbound passwords and
explicit capabilities
external accounts unique names and IDs
external identities
external passwords
F
filters dynamic
fine-grained controls
assigning
batch reporting
identity-driven properties
implementations of
ACT consolidation
baseline ACTs
examples
,
separated administration
G
granularity
unique names and IDs
groups compared with roles
identity precedence
management of
password updates for
PUBLIC
SAS Administrators
SASUSERS
E
algorithms
changing data in motion settings for
SAS servers
contexts for coverage
data at rest
data in motion
default settings for data at rest
default settings for data in motion
increasing strength for outbound passwords in transit
H
Hide baseline ACT
host access example
multiple levels of
to SAS tables
host authentication
I
ID formats
identity hierarchy
Index
279
identity passing
identity precedence
identity relationships network
identity-driven properties
identity, preserving
IDs, unique
implicit capabilities
import macros
import, user
See
information maps fine-grained controls and
inheritance paths
,
internal account policies changing
per-account
server-level
internal accounts unlocking
internal authentication
IWA (Integrated Windows authentication)
configuring
Kerberos protocol
K
Kerberos protocol
forcing
WriteMetadata and
WriteMemberMetadata permissions
metadata layer
metadata server authentication
determining SAS identity
align authentication for
SAS token authentication for
storing user IDs and passwords
N
names, unique
NETENCRALG system option
O
OBJECTSERVERPARMS system option
CEL parameter
Options tab workspace server
outbound domains
outbound passwords encryption for data in motion
increasing encryption strength for
L
launch credentials
creating and designating
criteria for
launching a standard workspace server
LimitData baseline ACT
lockout period
logging security events
logins authentication and
for web authentication
M
macros for user import and synchronization
ManageCredentialsMetadata permission
ManageMemberMetadata (MMM) permission
mediated access
metadata
P
PAM (pluggable authentication modules)
passwords
encryption strength for outbound passwords in transit
encryption strength for passwords at rest
enhanced protection for
external accounts
outbound, and encryption for data in motion
policies
SAS internal accounts
storing for mixed providers
storing for third-party servers
storing for workspace server
updates for users and groups
updates in the SAS Deployment
Manager
per-account policies
permission conditions
permissions
See also
assigning
by object
by task
280
Index
fine-grained controls
granularity and mechanics
identity relationships network
inheritance paths
,
key points for
ManageCredentialMetadata (MCM)
,
ManageMemberMetadata (MMM)
patterns of selected ACTs
relationship with capabilities
repository-level controls
resource relationships network
resource-level controls
setting
WriteMetadata (WM) and
WriteMemberMetadata (WMM)
pluggable authentication modules (PAM)
pooling
See
preservation of user identity
processing servers authentication to
Protect baseline ACT
PUBLIC access
PUBLIC group
R
reporting
See
repository-level controls
resource access opening up
resource relationships network
resource-level controls
RETURNPASSWORDS=
unique names and IDs
application features and
compared with groups
increasing or reducing availability of
management of
redistributing capabilities
S
SAS Administrators group
SAS identity determining
SAS internal authentication
SAS internal passwords
SAS OLAP Server fine-grained controls and
SAS Scalable Performance Data Server fine-grained controls and
SAS servers changing encryption settings for data in motion
SAS tables direct access versus mediated access
host access
risks of mediated host access
SAS token authentication
configuring
SAS.ExternalIdentity property
SAS.IdentityGroupName property
SAS.IdentityGroups property
SAS.IdentityName property
SAS.PersonName property
SAS.Userid property
SASProprietary encryption
SASUSERS group
scope of import process
of synchronization process
security
security checklist
metadata layer
passwords
security logging
%MDSECDS macro
authorization data sets
security tasks ensuring availability of application features
facilitating authentication
managing passwords
opening up access
server definitions access to
protecting
server-level accounts changing policies
benefits and risks of
servers preserving identity
service accounts password updates for
single sign-on (SSO)
,
standard workspace server authentication to
launching
synchronization
See
T
tables
Index
281
canonical
direct access versus mediated access
host access
risks of mediated host access
third-party servers authentication to
storing passwords for
,
configuring
trusted domains
trusted peer connections
trusted user connections
user synchronization
,
macros for
,
sample code
scope of
synchronizing identities
users adding regular users
dual users
ID formats
identity information
identity precedence
management of
password updates for
U
UNIX sample code for UNIX /etc/passwd import
unlocking internal accounts
user accounts identifying or creating
user definitions and
user administration
contents of
unique names and IDs
user identity preservation
user IDs storing for mixed providers
user import
,
canonical tables
external identities
importing identities
macros for
,
sample code for Active Directory import
sample code for generic file import
sample code for UNIX /etc/passwd import
scope of import process
W
web authentication
configuring
implementation instructions
logins for
Windows
IWA
workspace server authentication to
coordinating
credential gaps
launching
Options tab
storing passwords for
client-side
comparison of configurations
eligible requests actually using pooling
modifying the initial configuration
requests eligible for
server-side
WriteMemberMetadata (WMM) permission
WriteMetadata (WM) permission
282
Index
Laadi alla
Avalik link on uuendatud
Teie vestluse avalik link on uuendatud. You may view and manage your chat links after registration here.
Dokumendi avalik link
Reklaam
