HP OpenView OmniBack II Administrator`s Guide

HP OpenView OmniBack II
Administrator’s Guide
HP Part No. B1957-90021
Printed in USA
April 1996
Notice
Hewlett-Packard makes no warranty of any kind with regard to this material,
and specifically disclaims the implied warranties of merchantability and
fitness for a particular purpose. Hewlett-Packard shall not be liable for errors
contained herein or for any damages, whether direct, indirect consequential,
incidental, or special, in connection with the furnishing, performance, or use
of this material
The specific warranty information for the product as to which this material
related is available in Hewlett-Packard’s purchasing and licensing terms and
conditions.
© Copyright 1996 Hewlett-Packard Company
All rights are reserved. No part of this document may be photocopied
reproduced, or translated to another language without the prior written
consent of Hewlett-Packard Company. The information contained in this
document is subject to change without notice.
Restricted Rights Legend
Use, duplication, or disclosure by the Government is subject to restrictions as
set forth in subparagraph (c)(1)(ii) of the Rights in Technical Data and
Computer Software clause in DFARS 252.227-7013. Rights for non-DOD
U.S. Government Departments and Agencies are as set forth in FAR 52.22719(c)(1,2).
OSF/Motif is a trademark of the Open Software Foundation in the U.S. and
other countries.
HP-UX 9.x and 10.x for HP 9000 Series 700 and 800 computers are X/Open
Company UNIX 93 branded products.
X/Open is a registered trademark, and the X device is a trademark of X/Open
Company Ltd. in the UK and other countries.
UNIX® is a registered trademark in the U.S.A. and other countries, licensed
exclusively through X/Open Company Limited..
DOS, Windows, Windows for Workgroups, Windows 95, and Windows NT
are registered trademarks of Microsoft Corporation.
IRIX is a registered trademark of Silicon Graphics, Inc.
iFOR/LS is a trademark and Gradient is a registered trademark of Gradient
Technologies, Inc.
ii
Conventions
Conventions
To make the information easier to understand, this manual uses various
typographic conventions within the text, the following list explains these
conventions:
Computer Font
Computer font is used for the following:
• Directories and names:
/opt/omni/bin
• Screen prompts: Select (highlight) the device
with the Mount Request status.
• Messages displayed by the system: The
session is Pending then finally Running.
• What
you
type,
for
example:
Home_Directory
Bold
Bold face is used for command names, and the
names of windows and fields:
• Select Object → Add → Filesystem
• ..the OmniBack II Editor window appears
• ..in the Mountpoint field
OK
References to buttons on the display look like
this:
• Click on the Help button
• Click on the Options button
iii
In This Book
In This Book
HP OpenView OmniBack II Administrator’s Guide describes features and
functions of the HP OmniBack II network backup product.
Target Audience
This manual is intended for administrators who are responsible for the
planning, setup, and maintenance of the network backup.
The manual is organized as follows:
Chapter 1
An overview of the HP OpenView OmniBack II.
Chapter 2
Installation overview and tasks.
Chapter 3
An overview of concepts and tasks for configuration.
Chapter 4
How to run backups.
Chapter 5
How to restore files.
Chapter 6
Monitoring sessions
Chapter 7
Advanced tasks and concepts.
Chapter 8
Maintaining the OmniBack II Database.
Chapter 9
Using Complex Devices and Autochangers.
Chapter 10
Troubleshooting information.
Glossary
Definitions of terms related to the product.
iv
Contents
1. Introducing HP OpenView OmniBack II
Network Backups and HP OmniBack II .................................................... 1-3
OmniBack II architecture........................................................................... 1-4
OmniBack II components ....................................................................... 1-4
How a backup session works.................................................................. 1-6
How a restore session works................................................................... 1-7
Using the User Interfaces........................................................................... 1-9
Graphical user interface.......................................................................... 1-9
OmniBack II online help ...................................................................... 1-10
The command line interface ................................................................. 1-12
2. Installation Overview
Distributing HP OmniBack II to Clients.................................................... 2-3
Importing and Exporting Hosts.................................................................. 2-5
Importing hosts to the cell ...................................................................... 2-5
Exporting hosts from a cell..................................................................... 2-6
Updating OmniBack II Clients .................................................................. 2-8
To update OmniBack II clients ............................................................... 2-8
Deinstalling OmniBack II from a Host ...................................................... 2-9
To deinstall OmniBack II modules ......................................................... 2-9
3. Configuring a Cell
i
Contents
Planning the Cells .......................................................................................3-3
Planning Your Backup Strategy..................................................................3-5
Choosing the type of backup ...................................................................3-5
Choosing the mode of backup .................................................................3-6
Deciding what to back up ........................................................................3-9
Choosing a backup schedule ...................................................................3-9
Protection...............................................................................................3-10
Planning cell access...............................................................................3-11
Using data compression.........................................................................3-12
Using data encryption............................................................................3-12
Media Management ..................................................................................3-13
The Media Pool approach......................................................................3-13
Pool attributes........................................................................................3-19
Managing media compatibility..............................................................3-22
Media Identification ..............................................................................3-23
Configuring and Using Logical Devices...................................................3-25
The concept of logical devices ..............................................................3-25
Configuring logical devices...................................................................3-27
Configuring Typical Devices.................................................................3-30
Rescanning a Device .............................................................................3-35
Media Management Tasks ........................................................................3-37
Migrating from a Previous Media Pool .................................................3-37
Creating a media pool............................................................................3-38
Adding slots to an exchanger pool ........................................................3-38
Adding media to a pool .........................................................................3-39
Verifying a medium ...............................................................................3-40
ii
Contents
Erasing a medium ................................................................................. 3-40
Modifying a medium ............................................................................ 3-41
Viewing information about a medium .................................................. 3-41
Moving a medium................................................................................. 3-42
Recycling media/Removing protection ................................................ 3-42
Exporting a medium ............................................................................. 3-43
Importing a medium ............................................................................. 3-44
Scanning media..................................................................................... 3-45
Removing exchanger slots and media from the exchanger pool .......... 3-45
Entering media in an exchanger ........................................................... 3-46
Ejecting media from an exchanger ....................................................... 3-46
Viewing media scheduled for backup................................................... 3-47
Configuring User Classes and Users........................................................ 3-48
Configuring user classes ....................................................................... 3-49
Configuring users.................................................................................. 3-53
4. Backing Up Your System
Understanding Datalists ............................................................................. 4-3
Generating Datalists................................................................................ 4-4
Creating Datalists.................................................................................... 4-5
Modifying a datalist................................................................................ 4-6
Configuring a Filesystem Backup .............................................................. 4-7
Selecting afilesystem to backup.............................................................. 4-7
Selecting the filesystem backup options ............................................... 4-10
Configuring a Rawdisk Backup ............................................................... 4-11
iii
Contents
Specifying the rawdisk sections for backup ..........................................4-12
Selecting the rawdisk backup options ...................................................4-13
Running Interactive Backups....................................................................4-14
Starting a backup of an existing datalist................................................4-14
Status messages .....................................................................................4-16
Creating a temporary datalist ................................................................4-17
Previewing your backup ........................................................................4-18
Starting your backup .............................................................................4-18
Using Backup Options..............................................................................4-19
Description of backup options...............................................................4-22
Device Options ......................................................................................4-25
Transferring backup ownership .............................................................4-26
Setting Up Scheduled Backups.................................................................4-28
Setting up backup schedules..................................................................4-29
Skipping backups during holidays ........................................................4-34
Overriding scheduled backup options ...................................................4-34
5. Restoring Data
Restore Overview ......................................................................................5-3
Restoring Files from a Filesystem Backup.................................................5-4
Changing the view of backup objects......................................................5-4
Selecting the backup object .....................................................................5-5
Selecting the files to restore.....................................................................5-6
Selecting a File/Directory Version ..........................................................5-7
Specifying Fileset Restore Options .........................................................5-9
iv
Contents
Activating and monitoring the restore .................................................. 5-13
Searching for a specific file................................................................... 5-15
Restoring a Rawdisk Section (Logical Volume) ...................................... 5-17
Choosing the section to restore............................................................. 5-18
Specifying rawdisk restore options....................................................... 5-19
Starting and monitoring a rawdisk restore............................................ 5-19
Responding to mount requests.............................................................. 5-20
Restoring Files from a Rawdisk Backup ................................................. 5-21
Selecting the file to restore ................................................................... 5-22
Specifying rawdisk restore options....................................................... 5-23
Starting and monitoring file restore from a rawdisk............................. 5-24
Restoring a file when you do not know the filename............................ 5-25
6. Monitoring Sessions
Monitoring Running Sessions.................................................................... 6-3
Viewing details of a currently running session....................................... 6-3
Viewing Previous Sessions ........................................................................ 6-5
Viewing details on previous sessions ..................................................... 6-6
Monitoring Several Cells ........................................................................... 6-7
7. OmniBack II Advanced Tasks and Concepts
Automating Mount Prompt Handling ........................................................ 7-3
Mount prompt script examples ............................................................... 7-4
v
Contents
Using Pre- and Post- exec functions...........................................................7-6
Pre- and Post- exec command for a datalist ............................................7-6
Pre- and Post- exec commands for an object...........................................7-8
OmniBack II Global Options File.............................................................7-11
Managing Failed Backups ........................................................................7-12
Failed system backup ............................................................................7-12
Restarting failed backups ......................................................................7-13
Detection of CDF files ..............................................................................7-14
OmniBack II and Journaled File Systems on HP-UX 10.x ......................7-15
Concepts ................................................................................................7-15
Limitations.............................................................................................7-16
Restoring ...............................................................................................7-17
Script templates .....................................................................................7-17
Deleting a session .....................................................................................7-19
Recovery after a Crash..............................................................................7-20
Preparing for a crash..............................................................................7-20
Recovering a client system....................................................................7-21
Recovering a Cell Server.......................................................................7-22
Example of mission critical system recovery........................................7-23
Moving a Cell Server................................................................................7-26
Moving a Cell Server ............................................................................7-26
MC/ServiceGuard and OmniBack II ........................................................7-28
Functionality and Behavior ...................................................................7-28
vi
Contents
Prerequisites for Configuration............................................................. 7-29
Configuring Primary and Secondary Cell Servers................................ 7-29
Configuring of the Cell Server package................................................ 7-30
Starting the Cell Server......................................................................... 7-31
Backing up a Secondary Cell Server system ........................................ 7-32
Notification and Reporting Scripts........................................................... 7-33
The omninotify.ksh script............................................................... 7-34
The omnirpt.ksh script ..................................................................... 7-35
Customizing the Compression Algorithm ............................................... 7-37
Creating a compression code module ................................................... 7-38
Customizing the Encryption Algorithm................................................... 7-44
Creating an encryption code module .................................................... 7-44
8. Maintaining the OmniBack II Database
Installing and Configuring the Database.................................................... 8-3
Managing the Growth of the Database ...................................................... 8-4
Projecting the growth of the database..................................................... 8-4
Limiting the growth of the database ....................................................... 8-6
Purging obsolete data from the database ................................................ 8-7
Shrinking the size of the database file .................................................. 8-10
Backing Up the Database......................................................................... 8-12
Backing up the database using a datalist .............................................. 8-13
Restoring the Database ............................................................................ 8-14
vii
Contents
Recovering the Database .......................................................................8-15
Moving the Database to Another System .................................................8-17
9. Using Complex Devices and Exchangers
Managing Complex Devices with OmniBack II ........................................9-3
Complex device types .............................................................................9-3
How OmniBack II functions with complex devices................................9-4
Configuring a SCSI-II Exchanger...............................................................9-7
Creating an Exchanger Pool ....................................................................9-8
Selecting the device files .........................................................................9-8
Configuring the logical device ..............................................................9-11
Configuring a Pre 2.0 SCSI-II Exchanger ................................................9-13
Selecting the device files .......................................................................9-15
Configuring the logical device ..............................................................9-17
Configuring a Non-Standard Exchanger for Use with OmniBack II........9-20
Device Configuration ............................................................................9-20
Device Usage.........................................................................................9-22
10. Troubleshooting
Common GUI Errors and Warnings ........................................................10-3
Common Message Screen Errors and Warnings....................................10-10
Less Common Warning and Error Messages.........................................10-19
viii
1
Introducing HP OpenView OmniBack II
1-1
Introducing HP OpenView OmniBack II
In This Chapter
HP OpenView OmniBack II is the next generation of the OmniBack family
of backup and restore utilities. It expands the functionality of OmniBack and
OmniBack/Turbo by adding powerful new features that allow you to easily
manage large networks. These new features include a comprehensive new
media management system that tracks not only your backup media, but also
the data contained on those media. The flexibility provided by OmniBack II
enables you to develop customized backup policies while managing the
backup operations from a central point for each part of your organization.
This allows you to create a backup solution that better fits the needs of your
organization.
OmniBack II also enables the administrator to create a true "lights-out"
environment by automating the backup process using media management,
logical devices, and pre/post-processing functionality.
Lost or corrupted files create critical downtime until the files can be located
and restored. OmniBack II includes a powerful media management feature
that helps you to quickly locate and restore lost files or directories, while
guiding you through the process.
1-2
Introducing HP OpenView OmniBack II
Network Backups and HP OmniBack II
Network Backups and HP OmniBack II
Standard backup and recovery tools do not always fit the dynamic structure
of today’s organizations. HP OpenView OmniBack II provides a reliable,
flexible solution for your backup needs by introducing the concept of Cells.
Cells allow you to divide your network into independent logical units that
correspond to different backup strategies. HP OpenView OmniBack II lets
you manage these cells from a central point, such as initiating backup,
restoring, and monitoring actions on any cell in your network. This
flexibility provides a scalable backup solution that grows as your network
grows.
A cell can represent an organizational entity (such as a department) or a
geographical location (such as a field office). This allows you to design
highly-customized environments for the important functions of your
network. For example, a cell representing the payroll department can limit
backup and restore tasks to the administrator only, whereas a cell
representing the engineering department can allow any user to restore their
own files.
The complete HP OpenView OmniBack II structure and functionality is
integrated with HP OpenView OperationsCenter. This integration allows
you to manage networks consisting of a number of cells, each of which
contain many nodes. HP OpenView OperationsCenter lets you zoom in to
particular cells with its easy-to-use graphical display representing your
network.
For more information on using OmniBack II with OperationsCenter, see HP
OpenView OmniBack II Integration Guide.
1-3
Introducing HP OpenView OmniBack II
OmniBack II architecture
OmniBack II architecture
OmniBack II allows you to have many cells in the network, and each of
these cells operates as an independent part of the enterprise network. Each
cell is installed and configured separately. How the cells are configured
determines how backups are handled, which devices are used, and so forth.
Dividing the network into separate cells gives the network administrator
greater flexibility in developing a backup policy for each entity of their
organization.
A cell consists of a set of nodes; a Cell Server and Clients. One node can
only belong to one cell. The Cell Server is the main system in the cell and
runs the core OmniBack II software and also contains the OmniBack II
database. A client can be any system with the supported architecture.
Essentially, a client can be a system being backed up, or a system with a
backup device to which the backup data is stored, or both. The role of the
client depends on the software installed. See HP OpenView OmniBack II
Installation Guide and “Distributing HP OmniBack II to Clients” on
page 2-3 for installation instructions.
OmniBack II has a media management system which allows you to track the
backup media and the data contained on those media. The term media
refers to the physical media on which the backed-up files are stored. Media
exist in several forms: tape refers to DAT (Digital Audio Tape) or similar
magnetic tape; disk refers to a hard disk or rewritable optical disk.
OmniBack II components
OmniBack II consists of three basic components - User Interfaces, Cell
Server software, and Disk and Media Agents. Each of these components can
run on a different node in the cell. Figure 1-1 shows this hierarchical
architecture.
1-4
Introducing HP OpenView OmniBack II
OmniBack II architecture
On the first level, OmniBack II has two types of user interfaces, Graphical
User Interface (GUI), designed for simplicity of use, and the Command Line
Interface (CLI), designed for efficient use in scripts for experienced UNIX
administrators.
User Interface
Commands
Visual Cues
ols
ntr
ntr
Co
Co
ols
Sta
ntr
tus
Co
ols
tus
Figure 1-1
ntr
Disk Agent
Co
Sta
ols
Cell Server Software
Data Flow
Media Agent
HP OmniBack II Components
The second level is the Cell Server software, which runs on the Cell Server
node. When a command is entered on any node in the cell (using the GUI or
command-line), the following actions occur:
1. A connection (local or remote) to the Cell Server is established.
2. The Cell Server determines the type of command (backup, restore, or
status).
3. The Cell Server starts a Session Manager (SM) needed to execute the
command.
4. As the session progresses, the Session Manager writes information to the
OmniBack II database about the session and the data stored to the media
during backup.
The third level is the Client component which runs on the individual node.
The Session Manager (started in the second level) is responsible for
managing the clients needed to execute the command. To perform a backup
or restore, the Session Manager determines which devices are needed and
how these will be used. Next, it starts the processes needed to control these
devices. These processes are called Agents, and are either a Disk Agent
1-5
Introducing HP OpenView OmniBack II
OmniBack II architecture
(DA) or Media Agent (MA). The Disk Agent reads and writes data to a disk.
The Media Agent reads from and writes to a backup medium. Figure 1-2
shows how these agents operate during a backup or restore.
Backup Process
Disk Agent
Media Agent
Media
Disk
Restore Process
Media Agent
Disk Agent
Media
Figure 1-2
Disk
The Role of the Disk Agents and Media Agents
How a backup session works
A backup session starts when the backup request is initiated. During this
session, OmniBack II performs a backup of requested files and disks to the
requested media.
1. When the user initiates the backup, the Cell Server determines the type of
session that was requested (backup) and starts the appropriate Session
Manager.
2. The Session Manager reads the specified datalist and determines what
needs to be backed up and which devices will be used.
3. The Session Manager then determines how many Media Agents are
needed and starts the Media Agents and the Disk Agents needed for these
Media Agents. For example, a Media Agent is started for each media
(tape) drive that will be used and a Disk Agent is started for each disk that
will be read.
4. The Monitor window is displayed. This window lets you respond to
mount prompts and view the progress of a backup session.
1-6
Introducing HP OpenView OmniBack II
OmniBack II architecture
5. The Disk Agents and the Media Agent determine whether a local or
remote transfer is needed and the Disk Agents start sending data to the
Media Agent.
6. If more than one Disk Agent is used, the Disk Agents send data to the
Media Agents concurrently and the Media Agent interleaves the data on
the medium.
7. As each block of data is written to the media, the Media Agent sends
information to the Session Manager about what was backed up. The
Session Manager uses this information to update the catalog of backed up
files in the OmniBack II database.
How a restore session works
A restore session starts when the restore request is initiated. During this
session, OmniBack II performs a restore of requested files and disks from
the media.
1. When the user initiates the restore, the Cell Server determines the type of
session that was requested (restore), and starts the appropriate Session
Manager.
2. The Session Manager then determines which filesystems or directories
will be restored, which devices will be used, and what restore options
were specified.
3. The Session Manager starts the necessary Disk Agent and Media Agent.
For example, a Media Agent is started for the media (tape) drive that will
be used and a Disk Agent is started for the disk where the data will be
restored.
4. The Monitor window is displayed. This window lets you respond to
mount prompts and view the progress of a restore session.
5. The Disk Agent and the Media Agent determine whether a local or remote
transfer is needed and the Media Agent starts sending data to the Disk
Agent.
6. The Session Manager then updates the OmniBack II database.
1-7
Introducing HP OpenView OmniBack II
OmniBack II architecture
7. The Disk Agent writes the data to the disk.
1-8
Introducing HP OpenView OmniBack II
Using the User Interfaces
Using the User Interfaces
There are two user interfaces to OmniBack II: a Graphical User Interface
(GUI) and a command-line interface. Both can be accessed on the Cell
Console nodes, from where you can perform all OmniBack II tasks.
Figure 1-3
HP OmniBack II Main Menu
Graphical user interface
The graphical user interface uses colors and graphics to provide visual cues
to using OmniBack II features and functionality. It guides you through all
the administrative, backup, monitor and restore tasks.
This GUI uses X11/Motif features, such as buttons and list boxes to enhance
the usability. Whenever possible, list boxes are provided to allow you to
select from a list instead of typing in your selection. In addition, a
comprehensive online help system provides information about each window
and each task.
When in Monitor mode, the OmniBack II GUI uses color to alert you to
changes in the status of backup and restore sessions. You can monitor all
sessions in your cell, or focus on a specific session. The ability to monitor
and manage sessions is further enhanced with the addition of HP
OperationsCenter; each node is graphically represented on a map of the
network and changes color when status changes occur.
1-9
Introducing HP OpenView OmniBack II
Using the User Interfaces
The following naming convention is used to identify commands for the GUI
interface and the command-line interface; commands for the GUI start with
a word xomni followed by a word representing the function. For example,
the command xomnirestore starts the GUI for restore tasks only.
To use the graphical user interface of OmniBack II, enter:
xomni
to start the GUI for the complete OmniBack II
functionality
xomniadmin
to start the GUI for admin (configuration)
xomnibackup
to start the GUI for backup
xomnicellmon
to start the GUI for monitoring multiple cells
xomnimonitor
to start the GUI for monitoring a single cell
xomnirestore
to start the GUI for restore
The /usr/lib/X11/app-defaults/XOmni defines the appearance of the
graphical user interface and can be modified if desired.
OmniBack II online help
Information about OmniBack II is available from two sources: this manual
and the online help system. This manual contains the information you need
to plan and administer your OmniBack II network, and some more
commonly used tasks. The online help system contains the information you
need to perform all available tasks.
The online help system provides four types of help as shown here:
On Help
Provides help on using the OmniBack II help system.
On Window
Displays information on the purpose of each window and
how the fields of that window are used. Help on Window
is available for each window. Included is information on
the menu items, fields and buttons of the window.
On Tasks
Displays information for performing specific tasks, such
as scheduling a backup.
1-10
Introducing HP OpenView OmniBack II
Using the User Interfaces
On Version
Displays version and copyright information for
OmniBack II
You access the online help by either using the Help pull-down menu or Help
buttons provided on OmniBack II windows.
Figure 1-4
OmniBack II Online Help
Within the help text are hyperlinks (cross-references) to related information
and definitions. These hyperlinks help you navigate through the online help.
You click on the hyperlinked word or phrase to "jump" to the new topic.
Hyperlinked words and phrases are marked with either a solid or dashed
underline.
1-11
Introducing HP OpenView OmniBack II
Using the User Interfaces
The command line interface
The command line interface follows the standard UNIX format for
commands and options and provides complete OmniBack II functionality.
You can use these commands in scripts to speed-up your commonly
performed tasks.
The list provides a short description of each OmniBack II command. For
more details refer to the appropriate manpage for the command.
omniabort
Aborts a specified active backup or restore session.
omnib
Performs a backup of a specified datalist. All backup
options are supported.
omnicc
Checks and displays licensing summary for the cell.
omnicellinfo
Displays configuration information of the OmniBack II
cell, such as backup objects, media pools, devices, hosts
and the database.
omnicp
Copies data from one OmniBack II medium to another.
Devices used must be of the same device type. It only
runs locally on the system where both devices must be
located.
omnicreatedl
Automatically creates a datalist for the cell or the local
host.
omnidb
Queries the OmniBack II database for information on
backups and restores performed.
omnidbcheck
Checks and maintains the consistency of the OmniBack II
database. When this command is executed, backup or
restore sessions must not be running.
omnidbinit
Initializes the OmniBack II database. All information on
backup and restores sessions is lost.
omniminit
Initializes a medium. The medium is assigned to an
OmniBack II media pool and added to the database.
omnimlist
Displays the content of an OmniBack II medium. It scans
1-12
Introducing HP OpenView OmniBack II
Using the User Interfaces
the catalog of the medium and shows all objects and
sessions on the medium.
omnimm
Manages media and media pools, checks the protection of
a medium, maintains the contents of the repository of an
exchanger device.
omnimnt
Confirms or cancels the mount prompt request issued by a
logical device where no media is available for backup.
Once you have put the requested medium in the device,
you can confirm the mount request.
omnimver
Verifies the data on an OmniBack II medium. It reads the
data and checks for correct format. Optionally, CRC
checking of each block can be specified.
omnir
Performs a restore of OmniBack II backup. You can
restore any objects from the OmniBack II backup.
omnistat
Displays a status of an active backup or restore session.
You can display the status of all active sessions or just one
specified session.
omnisv.sh
Starts, stops, or displays the status of OmniBack II
processes.
omnitrig
Triggers the scheduled backups. This command is
normally executed in the cron script.
omnidownload
Displays the configuration of the device, as recorded in
the OmniBack II database. You can use this command to
display the configuration, or to create an ASCII file
containing the device configuration description, which
can then be modified and omniuploaded to the database.
omniupload
Uploads the device configuration from a file. You can use
this command to configure a new device, or to change the
configuration of the existing device by using the
omnidownload command first.
For more details on these commands see the appropriate man pages.
1-13
Introducing HP OpenView OmniBack II
Using the User Interfaces
1-14
2
Installation Overview
2-1
Installation Overview
In This Chapter
OmniBack II introduces the concept of cells. By dividing a network into
cells representing the entities of an organization, the network administrator
can define a unique backup strategy for an organizational entity. How you
organize your network determines the backup efficiency of your cell. Each
cell is installed and configured separately.
Figure 2-1
The OmniBack II Cell Administration Window
Installing OmniBack II consists of the following tasks:
•
Planning the cells. See Chapter 3, “Configuring a Cell,” on page 3-1 for
more information.
•
Installing the OmniBack II software package on the Cell Server. See HP
OpenView OmniBack II Installation Guide for more information.
•
Distributing the OmniBack II modules on the clients in the cell. See
“Distributing HP OmniBack II to Clients” on page 2-3 for more
information.
•
Configuring the devices, users, and media for the cell. See the appropriate
sections in Chapter 3, “Configuring a Cell,” on page 3-1 for more
information.
2-2
Installation Overview
Distributing HP OmniBack II to Clients
Distributing HP OmniBack II to Clients
Once you have installed the complete OmniBack II software on the Cell
Server, you must distribute the appropriate OmniBack II software modules
to all hosts in the cell using the GUI. The function of a host is determined by
the software modules you install. The following list shows these modules:
Cell Console
A user interface to OmniBack II. It must be installed to all
hosts from which you want to control OmniBack II.
Disk Agent
Agent that reads and writes from a disk containing data. It
must be installed on all hosts which you want to back up.
Media Agent
Agent that reads and writes from a device containing a
backup medium. It must be installed on all hosts where
the backup devices are configured.
See the HP OpenView OmniBack II Integrations Guide for more information
on the following integrations.
OpC
Integration software for OperationsCenter integration.
OmniStorage
Integration software for OmniStorage filesystems backup.
SAP
Integration software for SAP database.
Oracle
Integration software for Oracle database.
Sybase
Integration software for Sybase database.
To simplify installation, your installation plans should include a list of hosts,
grouped by modules to be installed. For example, hosts 1,2, and 3 have the
Media Agent module, hosts 4,5, and 6 have the Media Agent and the Disk
Agent. You install hosts 1,2, and 3 as one group, and hosts 4,5, and 6 as
another.
2-3
Installation Overview
Distributing HP OmniBack II to Clients
NOTE:
To perform the installation, you will need either root access or an account with
root capabilities for every node on which you want to install HP OmniBack II
modules.
1. At the OmniBack II Main Menu or Admin Menu select Install to open
the Cell Administration window.
2. Select Edit → Install hosts to open the Cell Administration Install window.
3. Select the Add... button to open the Install - Load Hosts window.
The Domain field of the window shows the IP address for your cell. To
change this IP address, click on the field with the address and enter the
new IP address.
You can select hosts by typing their names in the Filter field and
selecting Add. Short and long hostnames can be used. You can also use
wildcards to select several hosts at once.
You can select hosts by their names or their IP addresses by entering the
exact host IP address or a range of addresses to add several hosts at once.
4. When you have specified all hosts that will have the same modules, close
the Install- Load Hosts window and return to the Cell
Administration - Install window. The hosts you specified now
appear in the Installation Target Hosts field.
5. Select which OmniBack II modules you want to install on selected hosts.
6. To confirm this host configuration, click the OK button. Follow
instructions and messages in the HP OmniBack II - Install
Terminal window.
2-4
Installation Overview
Importing and Exporting Hosts
Importing and Exporting Hosts
When you distribute software from the Cell Server to other hosts, they are
automatically added to the cell. If you need to modify your cell
configuration, move the host from one cell to another, or perform similar
tasks without removing the OmniBack II software from the host, you can
export and import hosts.
The manual importing of the host is useful, for example, when you want to
do one of the following:
•
•
•
•
•
An emergency fix for problems caused by insufficient licenses. By
exporting a host from the cell, its license becomes available to some other
host. This is, however, only a temporary solution until you order and get
more licenses.
Add a pre-installed host to the cell.
Move the host from one cell to another. You do not need to de-install
software from the host before moving it to another cell.
Remove from the configuration any hosts which are no longer a part of the
network.
Needed to integrate Novell, Windows NT, and Windows 95 clients. See
HP OpenView OmniBack II Integrations Guide for instructions.
Importing hosts to the cell
Importing a host tells the Cell Server to add the host (where OmniBack II
Modules are already installed) to the cell.
To add a host, follow these steps:
1. In the HP OmniBack II - Cell Administration window select
Edit → Import… from the menu bar to open the Cell
Administration - Import Hosts window.
2-5
Installation Overview
Importing and Exporting Hosts
2. Select Add... to open the Import - Load Hosts window.
The Domain field of the window shows the IP address for your cell. To
change this IP address, click on the field with the address and enter the
new IP address.
You can select the hosts by typing their names in the Filter field and
selecting Add. Short and long hostnames can be used. You can also use
wildcards to select several hosts at once.
You can select hosts by their names or their IP addresses by entering the
exact host IP address or a range of addresses to add several hosts at once.
3. When you have specified all the hosts that you want to import, close the
Import - Load Hosts window and return to the Cell
Administration - Import Hosts window. The hosts you specified
now appear in the Hosts field.
4. Click the Import button to import the hosts.
Exporting hosts from a cell
Exporting a host tells the Cell Server to remove the host from this cell
without removing the OmniBack II modules.
To export a host, follow these steps:
1. In the HP OmniBack II - Cell Administration window select
hosts you want to export from the cell.
2. Select Edit → Export from the menu bar. The selected hosts are
displayed in the Hosts field.
3. You can remove the hosts from the list by selecting them and clicking
Delete.
4. Click the Export button to export the hosts in the Hosts list.
5. Remove all appearances of the exported hosts from datalists. See the
online help for instructions how to modify the datalists.
2-6
Installation Overview
Importing and Exporting Hosts
NOTE:
If you do not remove the exported hosts from datalists, OmniBack II will try to back
up unknown hosts and this part of the datalist will fail with the following error
message:
[Critical] From: BSM@finance.hq ”” Time: 09/17/94 14:01:07 Host(s)
Backup session participant(s) not in confirmed configuration
2-7
Installation Overview
Updating OmniBack II Clients
Updating OmniBack II Clients
If you want to update to a new version of OmniBack II, you can do it easily
by installing the new software on the Cell Server and then performing this
procedure to update all hosts in the cell. You can manually select which
modules you want to update. OmniBack II automatically checks all selected
systems in the cell and updates the selected modules.
OmniBack II automatically removes installation packages of the older
version thus giving more disk space.
This feature is supported for UNIX systems only.
To update OmniBack II clients
This procedure does not install any new modules, it just updates existing
ones.
It may take some time, depending on your cell configuration.
1. In the HP OmniBack II - Cell Administration window, select
Edit → Update All to open the Cell Administration - Update
Hosts window.
2. Select the modules you want to update and start the update by clicking
OK.
3. Follow instructions and messages in the HP OmniBack II - Update
Terminal window.
The new version number of the modules you updated now appears for all the
hosts in the HP OmniBack II - Cell Administration window.
2-8
Installation Overview
Deinstalling OmniBack II from a Host
Deinstalling OmniBack II from a Host
If your system configuration changes, you may need to deinstall OmniBack
II software from the host. Deinstalling OmniBack II from a host removes all
the OmniBack II modules from the host and removes all references to this
host from the Cell Server.
This feature is supported for UNIX systems only.
To deinstall OmniBack II modules
1. In the HP OmniBack II - Cell Administration window select the
host you want to deinstall.
2. Select Edit → Deinstall from the menu bar to open the Cell
Administration - Deinstall window.
3. Select modules you want to deinstall. Only the modules currently installed
on this host are available for deinstallation. If you deinstall all the
available modules, this host is completely removed from the OmniBack II
cell.
4. Click OK to begin the deinstallation process. Follow instructions and
messages in the HP Cell Administration - Deinstall
Terminal window.
5. Remove all references to this host from datalists. See the online help for
instructions how to modify the datalists.
NOTE:
If you do not remove the host from your datalists, OmniBack II will try to back up an
unknown host and this part of the datalist will fail with the following error message:
[Critical] From: BSM@finance.hq ”” Time: 09/17/94 14:01:07
Host(s)
Backup session participant(s) not in confirmed configuration
2-9
Installation Overview
Deinstalling OmniBack II from a Host
2-10
3
Configuring a Cell
3-1
Configuring a Cell
In This Chapter
Configuring a cell involves defining a backup strategy, configuring your
logical devices, implementing a media management strategy, and setting up
the user access security. The information you need to plan your cell
configuration is explained in the following sections. The actual steps for
doing the configuration are explained in the online help system. The online
help system is divided into task help and window help. The task help
provides detailed step-by-step information for configuring a cell. The
window help provides detailed information about the fields and buttons for
each window.
3-2
Configuring a Cell
Planning the Cells
Planning the Cells
A cell can represent an organizational entity (such as a department) or a
geographical location (such as a field office). It consists of a set of nodes
(server and clients) having a common backup policy, located in the same
time zone, and existing on the same LAN. How the network is divided into
cells has a major impact on the backup strategy for the network, so careful
planning is important. For the best performance, we recommend having a
maximum of 50 nodes in the cell. Additional factors that affect performance
include the amount of data and the number of files to be backed up.
NOTE:
A node can belong only to one cell.
HP OmniBack II allows you to manage and monitor the backup and restore
activities at either the cell or enterprise level. An administrator defines the
backup policies, security, logical devices, media pools, and users for each
cell.
The HP OperationsCenter enhances your ability to manage and monitor the
backup and restore activities of all cells simultaneously at the enterprise
level. This is useful for very large networks.
When defining the cells, consider the following factors:
Quantity of
Data
Type of Data
Importance of
Data
The amount of data to be backed up.
Files, databases, specific applications (see HP OpenView
OmniBack II Integration’s Guide)
Is data modified often and has to be backed up regularly,
such as user’s working files, or is mostly permanent, such
as system files ?
Backup
Operations
Are unattended backups needed ?
Devices
Availability and characteristics of backup devices
3-3
Configuring a Cell
Planning the Cells
Security
User access to backup and restore functions
Performance
Network infrastructure and speed of the network
Time Zones
Calculations of backup start times
To install OmniBack II, you will need to define the following for each cell:
•
Which node will be the Cell Server (Note: once the Cell Server is selected,
it is difficult to change it)
•
Which nodes have disks to be backed up (Disk Agent clients)
•
Which nodes have backup devices that will be used (Media Agent clients)
•
Which nodes (in addition to the Cell Server node) will be used to access
OmniBack II (command-line interface and the Cell Console)
Once you have planned your cells, you can begin installing the HP
OmniBack II on the Cell Server. See HP OpenView OmniBack II Installation
Guide for instructions.
3-4
Configuring a Cell
Planning Your Backup Strategy
Planning Your Backup Strategy
OmniBack II offers great flexibility in how you configure and perform your
backups. When you plan your backup strategy, you need to make the
following decisions:
backup type
Filesystem or Rawdisk
backup mode
Full backup or Incremental backup
backup contents Which data to back up
backup options Compression, encryption, public/private access to data,
report level, file locking, date/time stamps, pre/post
execution of commands, and data protection limits
backup
schedule
Specific date/times or time cycle
Each of these is explained in the following sections.
Choosing the type of backup
How your data is organized influences the type of backup you choose. Two
types of backup are explained here:
Filesystem
If you want a complete record of every file and directory that was backed up,
a Filesystem backup is best. This type of backup tracks the logical structure
of the data on the media. The advantage is easier recovery of data; the
disadvantage is that filesystem backups are generally slower than rawdisk
backups. This backup type is used for ordinary filesystems such as system
and user directories and so forth.
3-5
Configuring a Cell
Planning Your Backup Strategy
Rawdisk
Rawdisk backups are useful, for example, for backing up databases located
on a disk without a file system. This type of backup does not track the
logical structure of the data on the media. The advantage is faster backup of
a complete disk. The disadvantage is that the OmniBack II does not know
which files are part of the backup when you do a restore. Only the name of
the rawdisk device file is recorded in the database. The rawdisk backup
backs up a complete disk, including the unused parts and may therefore take
longer than a filesystem backup when a disk is only half full.
Note that raw disk backup is only supported for disks smaller than 2 Gb.
Choosing the mode of backup
How frequently you back up your data affects the security and recoverability
of the data. The more frequent the backup, the smaller the potential data loss
in the event of a disk failure. There are two types of backup mode: Full and
Incremental.
Full Backups
A full backup includes all the data for the specified objects (directories,
filesystems, or databases). It is a "snapshot" of the data as it currently exists.
In the event of data loss, all data can be recovered using this one backup.
The advantage of a full backup is security (the complete object is backed up)
and faster, easier recovery (only one set of backup media must be restored).
The disadvantage is that a backup may take long and needs more media to
back up all data.
Incremental Backups
An incremental backup includes only the data that has been modified since
the last full or last incremental backup. Incremental backups are faster and
need less media to back up data, but makes the restore of all data more
complicated as you usually need all media used since the last full backup.
Incremental backups are only valid if the full, protected backup on which
they are based exists. See “Using Backup Options” on page 4-19 for more
information. OmniBack II checks if there is a protected full backup before
3-6
Configuring a Cell
Planning Your Backup Strategy
starting an incremental backup. If there is no protected full backup,
OmniBack II performs a full backup instead. This can potentially overwrite
your unprotected full backup. You can customize this behavior by setting the
IncrOnProtected variable in the global options file. See “OmniBack II
Global Options File” on page 7-11 for more information.
Incr
Incr (1-9)
Is based on the last backup, either full or incremental.
This is a basic incremental backup type.
Are based on the previous incremental backup of a lower
level. For example, incremental backup level 3 backs up
data changed from the last level 2 or lower backup.
Three examples of incremental backups (incremental, single-level and
multiple-level) are described here:
Incremental Backups
Using incremental backup (Incr) provides an efficient and simple way to
configure backups. Each incremental backup is based on the previous
backup of any type.
Single Level Incremental Backups
A typical backup strategy involves a combination of full and incremental
backups. Each incremental backup is based on the last full backup. For
example, a full backup is done once a week during a time of low system
usage and incremental backups are done at once each business day in the
evening. This type of schedule has the advantage of minimizing the time
needed for recovery while at the same time minimizing the effort needed to
back up the data on a regular basis. Here is an example of this type of
schedule:
Friday night
Full Backup includes all data
Monday night
Incr-1 Backup which includes all data modified since the
full backup on Friday night
Tuesday night
Incr-1 Backup which includes all data modified since the
full backup on Friday night
Wednesday
night
Incr-1 Backup which includes all data modified since the
full backup on Friday night
3-7
Configuring a Cell
Planning Your Backup Strategy
Thursday night Incr-1 Backup which includes all data modified since the
full backup on Friday night
In this example, a complete recovery on Wednesday involves restoring the
full backup from Friday and the level 1 incremental backup from Tuesday.
Everything between the backup of Tuesday night and the time of the system
disaster on Wednesday is at risk. Note that each incremental backup makes
the previous one obsolete.
Multiple Level Incremental Backups
Another backup strategy involves the use of multiple-level incremental
backups . Here, the incremental backups are based on other incremental
backups. By using different incremental levels, you can develop a
sophisticated backup strategy. Multiple-level incremental backups have the
advantage of reducing the time and number of media needed to perform the
incremental backups even more than single level incremental backup. The
disadvantage is in the increased complexity and length of time needed for
data recovery.
Here is an example of this type of schedule:
Friday night
Full Backup includes all data
Monday night
Incr-1 backup which includes all data modified since the
last full backup on Friday night
Tuesday night
Incr-2 backup which includes all data modified since the
Incr-1 backup on Monday night
Wednesday
night
Incr-1 backup which includes all data modified since the
full backup on Friday night
Thursday night Incr-2 backup which includes all data modified since the
Incr-1 backup on Wednesday night
In this example, a complete recovery on Wednesday involves restoring the
full backup from Friday, the level 1 incremental backup from Monday, and
the level 2 incremental backup from Tuesday. Everything between the
backup of Tuesday night and the time of the system disaster on Wednesday
is at risk.
3-8
Configuring a Cell
Planning Your Backup Strategy
Deciding what to back up
The previous two sections covered choosing a backup type and backup
mode. A third factor affects the security and recoverability of your data:
deciding WHAT to back up. Your system probably has a mixture of data,
some of which is very dynamic and some of which is static. Using the
example of a purchasing department, the vendor information is probably
static, but order information is very dynamic. The purchasing department
has a group of vendors from whom orders are placed on a continuous basis.
Assuming that there is a vendor database and an order database, the vendor
database remains somewhat static and the order database is updated on an
hourly basis. Therefore, the potential data loss is greater for the order
database because there are more database updates in the time between the
backups. A backup strategy for this scenario could include daily full
backups of the order database and daily incremental backups of the vendor
database.
Choosing a backup schedule
The previous three sections covered deciding WHAT and HOW to back up
data. This section helps you decide WHEN to back it up. For backups to be
effective, they must be done on a regular basis. The security of your system
is only as safe as your backups. The greater the time interval since your last
backup, the greater risk to your data. Backup schedules are generally set in
one of two ways: by specific date/time or by time cycle.
Specific Date/Time
The most common form of backup schedule is based on specific dates/times.
A backup is performed at the same time and on the same day each week or
month. For example, a full backup is done every Friday night at 6:00 and an
incremental backup is done every Monday, Tuesday, Wednesday, and
Thursday nights at 8:00. A schedule like this is easy to plan and remember.
OmniBack II can also be configured to check the Unix holiday file to skip
backups on holidays.
3-9
Configuring a Cell
Planning Your Backup Strategy
Time Cycle
Time cycle backups are done on regular time intervals based on periodic
schedules. For example, a full backup is done once every 2 days at 12:00.
Therefore, if the first backup is Monday at 12:00, subsequent backups will
start at 12:00 on Wednesday, Friday, Sunday, Tuesday, Thursday, etc.
Protection
This option enables you to set periods of protection for data you back up.
This prevents backup data from being overwritten for a specified period. The
default value is Permanent.
The backup protections are:
None
CAUTION:
Provides no protection. Data will be overwritten if the
medium it is stored on is reused.
Use this option for test backup only. Stored data will be overwritten when the
medium is used again since no protection has been chosen.
Until
Data on the medium cannot be overwritten until a
specified date. You enter the year, month, and day.
Protection for the data will stop at noon of the specified
day.
Days
Data on the medium cannot be overwritten for the number
of days specified. For example, if you set the protection
for 2 days and the current backup starts at 20:00 (8:00
PM) on Wednesday, the data will be protected until Friday
at 20:00.
Weeks
Data on the medium cannot be overwritten for the number
of weeks specified. For example, if you set the protection
for 1 week and the current backup starts at 10:00 on
Wednesday, the data will be protected until 10:00 the
following Wednesday.
Permanent
Data is permanently protected from being overwritten.
Use the recycle feature to remove the permanent
3-10
Configuring a Cell
Planning Your Backup Strategy
protection.
Planning cell access
OmniBack II allows you to run backup and restore sessions on a cell from
any Cell Server, as long as you have network access. This is particularly
useful when you manage more than one cell in your environment, you run
the HP OpenView OperationsCenter, or you have configured an additional
Cell Server with a switch over.
In smaller environments where you only have one cell, you may want to
disable this feature and allow starting backups from the local Cell Server
only.
The following actions alter the /usr/adm/inetd.sec (on HP-UX 9.0x) or
/var/adm/inted.sec (on HP-UX 10.x) file.
Limiting Access
Once you have configured all hosts in a cell, you must do the following to
limit access to the cell:
1. From the HP OmniBack II Install Window choose Edit → Add
access limit to cell to open the Cell Administration - Limit Host
Access window.
The Cell Server field displays the name of the local Cell Server and
cannot be modified.
In the OPC hosts(s) field enter hostnames of the HP OperationsCenter
systems from which you want to access systems in the OmniBack II cell.
In the Alt. cell server(s) field enter hostnames of alternate Cell
Servers.
2. Choose OK to limit access to the cell. The Messages field shows the
progress.
If you wish to add a new host later, you must repeat the procedure.
3-11
Configuring a Cell
Planning Your Backup Strategy
Allowing Access to the Cell
If you want to allow access to the cell from other Cell Servers or you run the
HP OpenView OperationsCenter, you must:
1. From the HP OmniBack II Install Window choose Edit → Remove
access limit from cell to open the Remove access limit window.
2. Choose Remove limit to allow access to the cell from other Cell Servers.
Using data compression
OmniBack II has a compression feature which allows you to reduce the
number of media needed for a backup. The standard Unix compression
algorithm is used, which provides a compression ratio of approximately
50%. You can substitute a customized compression algorithm by replacing a
shared library and integrating this into OmniBack II. Compression can be set
for each object or for the whole session. For information on customizing the
compression algorithm, see “Customizing the Compression Algorithm” on
page 7-37.
NOTE:
Most modern backup devices provide built-in hardware compression that you can
configure when adding the device to the system. See your device’s manual for
instructions. In this case do not use the compress option of OmniBack II, since double
compression only decreases performance without giving better compression results.
Using data encryption
OmniBack II has a simple encryption feature which provides greater
security for the data contained on your backup media. Encryption can be set
at the object or session level. You can substitute a customized encryption
algorithm by replacing a shared library and integrating this into OmniBack
II. For more information on customizing the encryption algorithm, see
“Customizing the Encryption Algorithm” on page 7-44.
3-12
Configuring a Cell
Media Management
Media Management
OmniBack II introduces a complete media management system. It protects
your data by providing the following features:
•
•
•
•
protection against accidental overwrite
notification of bad or worn media
optimization of the space available in exchangers
tracking and managing media
OmniBack II provides protection by tracking and protecting the data on the
media. The media management system checks the media management
database before writing to a backup medium. This prevents accidental
overwrites.
The media management system uses a database to record information about
the following aspects of a backup or restore:
•
•
•
•
•
•
NOTE:
date and time the backup was started
type of backup
directories, files and disks that were backed up
file positions on the media
usage and condition of the media in the pools
protection expiration date for the data contained on the backup media
The growth and size of the media management database are determined by the level
of detail recorded about each backup and restore session. For more information on
maintaining the media management database, see Chapter 8, “Maintaining the
OmniBack II Database,” on page 8-1.
The Media Pool approach
The OmniBack II media management paradigm of handling backup
media uses the term ‘media pool’ to introduce a mechanism for managing
large numbers of media cartridges.
3-13
Configuring a Cell
Media Management
A media pool is an organized collection of media cartridges which share
common properties. Grouping media into media pools allows OmniBack II
to optimize media usage, allocation and retention policies. Each medium
belongs to only one pool at a time. The administrator specifies a particular
allocation and protection policy for media in this pool. OmniBack II also
tracks the amount of free space remaining in the pool based on the amount of
space specified for each medium when added it to the pool. OmniBack II
sums up the space on each medium to calculate the space for one pool.
In order to provide support for different types of backup devices,
OmniBack II provides different types of pools. Each of these is used with
different types of backup devices, but they all share the basic definition they define collections of related media which share common properties.
OmniBack II media management uses a pool-oriented approach; instead of
focusing on backup drives, we suggest the user first think about backup
media and organize this media into meaningful and managable pools.
Logical devices are then defined which ‘use’ these pools.
Benefits of this pool-oriented approach can be seen especially when we start
configuring complex backup devices, such as multi-drive SCSI II
exchangers or large medium libraries with embedded media management
functionality such as StorageTek ACS or Grau ABBA devices.
Standard pool
This type of pool is provided for devices such as:
•
Standalone tape (or optical) drives
•
Simple stacker devices
•
External control exchanger
•
Optical jukeboxes (not configured as SCSI exchangers)
In previous releases of OmniBack II this pool was also used to handle SCSI
exchangers. This method is now obsolete, and SCSI exchangers should be
configured to use the exchanger pool described below.
3-14
Configuring a Cell
Media Management
A standard pool is an organized set of OmniBack II media with the
following properties:
•
All media are initialized and owned by OmniBack II exclusively.
•
All media are of the same type, for example, 4mm tapes, MO disks, and
so forth.
•
All media within one pool share a common media allocation and usage
policy.
•
Physical handling of these media is typically not controlled by
OmniBack II.
The following actions can be performed on a standard pool and the media in
the pool:
“Adding media to a pool” on page 3-39
“Modifying a medium” on page 3-41
“Verifying a medium” on page 3-40
“Scanning media” on page 3-45
“Viewing information about a medium” on page 3-41
“Moving a medium” on page 3-42
“Recycling media/Removing protection” on page 3-42
“Erasing a medium” on page 3-40
“Exporting a medium” on page 3-43
“Importing a medium” on page 3-44
Magazine pool
Magazine pools are provided to allow comfortable and efficient handling
of small SCSI exchangers which allow their complete repository (set of
cartridges accessable to the exchanger’s robotics) to be replaced at once.
OmniBack II treats the whole magazine as one unit: a magazine will be
managed and handled completely as a unit and individual cartridges will not
be replaced. A magazine pool can contain a number of magazines. In
particular, this is true for the HP SoreStore 12000e exchanger which uses
magazines of six 4mm tapes that can be replaced with one operation.
3-15
Configuring a Cell
Media Management
A magazine pool can only handle OmniBack II owned media. If media is
exported, one can no longer control this media using OmniBack II anymore,
until this media is imported again. The operations on each medium in the
magazine are completely controlled by OmniBack II.
OmniBack II remembers in which magazine media resides. If the media in
the magazine are mixed up, OmniBack will issue a mount request for the
best available magazine.
OmniBack II automatically initializes magazines and the media in the
magazine if they are new. You may want to initialize magazines and media
before backup to save backup time.
The Prealloc Media backup option is ignored with magazine pools. See
“Description of backup options” on page 4-22 for more information.
During restore, OmniBack loads the medium from the slot where it should
be and issues a mount request if the medium is not found.
Magazine pools can be shared between multiple logical devices as long as
the devices have the same size magazine.
Cleaning Tape
Additionally, the magazine pool supports automatic cleaning of an
exchanger with cleaning tape configured. With cleaning tape support
enabled, the following is true:
•
OmniBack II expects the cleaning tape to be located in the last slot of the
magazine (for example, slot 6 on the HP SureStore 12000e exchanger.)
•
OmniBack will not use slot 6 for backup even if it contains a backup tape
and not a cleaning tape.
•
When dirty drive is detected, OmniBack II automatically cleans it using
the cleaning tape from slot 6.
See “The concept of logical devices” on page 3-25 for more information on
how to enable cleaning tape support.
3-16
Configuring a Cell
Media Management
The following actions can be performed on a magazine pool or the media in
the pool:
“Adding media to a pool” on page 3-39
“Modifying a medium” on page 3-41
“Verifying a medium” on page 3-40
“Scanning media” on page 3-45
“Viewing information about a medium” on page 3-41
“Moving a medium” on page 3-42
“Recycling media/Removing protection” on page 3-42
“Exporting a medium” on page 3-43
“Importing a medium” on page 3-44
Exchanger pool
This pool is provided for efficient management of large SCSI II exchangers,
such as HP MO, ADIC, SpectraLogic STL, ATL 4152, ExaByte, and so
forth (see HP OpenView OmniBack II Software Release Notes for a detailed
list of supported devices.) For more information on how to configure the
exchangers, see Chapter 9, “Using Complex Devices and Exchangers,”.
In addition to providing OmniBack II media management method to the use
of media pools, it also provides a better overview and handling of media.
The benefit of using the exchanger pool, is that the repository can be shared
between logical devices of a multi-device exchanger. The exchanger’s
repository may contain either OmniBack II or non-OmniBack II owned
media. The exchanger pool can handle media that are physically located in
the exchanger’s repository (resident)as well as media located somewhere
else (non-resident), for example, for safety reasons. This becomes
clearer when we think of configuring larger exchangers that are also used by
other applications, not just OmniBack II, and where media are often
removed from the repository to store it in a safe place. The exchanger pool
provides the possibility of sharing the repository with multiple drive
exchangers.
The exchanger pool allows you to also have full control of non-OmniBack II
owned media used by the exchanger, as long as they use barcodes and can
therefore be tracked. The OB2BARCODE=1 variable in the
/usr/omni/.omnirc (on HP-UX 9.0x) or /opt/omni/.omnirc (on HP3-17
Configuring a Cell
Media Management
UX 10.x) file on the system where the Media Agent for the device is running
must be set to allow OmniBack to use barcodes. See HP OpenView
OmniBack II Software Release Notes for a list of devices that have barcodes
configured by default.
CAUTION:
An exchanger pool can only be configured for a single exchanger - do not
use the same exchanger pool for different exchanger devices!
However, a single pool can be used for all drives in the exchanger.
Each drive of the exchanger must be configured as a separate OmniBack II
logical device.
The following actions can be performed on an exchanger or its media:
“Adding slots to an exchanger pool” on page 3-38
“Removing exchanger slots and media from the exchanger pool” on
page 3-45
“Entering media in an exchanger” on page 3-46
“Ejecting media from an exchanger” on page 3-46
“Adding media to a pool” on page 3-39
“Modifying a medium” on page 3-41
“Verifying a medium” on page 3-40
“Scanning media” on page 3-45
“Viewing information about a medium” on page 3-41
“Moving a medium” on page 3-42
“Recycling media/Removing protection” on page 3-42
“Erasing a medium” on page 3-40
“Exporting a medium” on page 3-43
“Importing a medium” on page 3-44
3-18
Configuring a Cell
Media Management
Pool attributes
This section provides a list of all attributes of different pool types. Note that
some attributes are specific and do not appear with all types of pools.
Pool Name
The pool name identifies the pool. It can consist of up to 32 characters
including spaces. We recommend that you assign a name which will help
you identify either the logical device or the function of that pool.
Media Class
Media class is the type of media, for example, DAT tapes or optical disks.
The Media Class button of the Pool Editor window has a menu of
supported media classes.
When you select the type of media, OmniBack II calculates available space
for that pool based on the media class you have selected. For example, if you
specify DAT, a default size of 2.0 GB is assumed when initializing the
media. You can overwrite this default by your own value.
Media Allocation Policy
The media allocation policy determines the order in which media are
accessed within a pool so that media wear out evenly. OmniBack II balances
the use of the media in a pool and therefore issues mount prompts based on
the status (condition) of the media.
Strict
directs OmniBack II to require a specific medium.
Medium must be already initialized. This allocation
policy is preferrable for use with exchanger devices to
prevent accidental overwrite of non-OmniBack II media
3-19
Configuring a Cell
Media Management
in the exchanger.
Loose
NOTE:
directs OmniBack II to accept any suitable medium in the
pool (medium must not be in a poor condition or
protected.) A non-OmniBack II medium is automatically
initialized. This policy is preferrable for stacker devices,
as they load media in sequential order - a strict policy
would require media to be loaded in the same order as
would be used.
For exchangers containing several media, a strict policy directs OmniBack II to issue
mount prompts for the least-used media first. It continues to request these "newer"
media until the usage of these media equals the usage of the other media in the pool.
Therefore, if you add a new (unused) medium to this pool, it will be requested more
often by OmniBack II. To balance the usage, we recommend that you pre-initialize
your media before adding them to an existing pool. Initializing a medium also allows
you to add a label to the medium, helping you to identify it.
Media Usage Policy
The media usage policy determines whether or not a new backup can be
appended to a medium containing a previous backup.
Appendable
NonAppendable
Appendable of
incrementals
only
3-20
A backup session writes data to the space remaining on
the last medium used in the previous backup session.
Subsequent media in the backup set are used from the first
position on each medium. Appending media conserves
media space but reduces flexibility in manipulating
media, because one medium can contain data from several
backup sessions.
A backup session writes data beginning at the first
position on the first medium of the backup media set.
A backup session appends to a tape only if an incremental
backup is performed. This allows having a complete
Configuring a Cell
Media Management
backup of full and incrementals on the same media if
there is enough space.
Media Condition Factors
OmniBack II calculates the condition of used media using media condition
factors. The condition of the media in a pool determines the condition of the
pool. The condition of a pool indicates the reliability of that pool for
backups. For example, old or worn media are more likely to have read/write
errors. OmniBack II allows you to define the acceptable thresholds for your
media. There are three types of conditions for a medium:
Good
less than 80% of the threshold for age or usage
Fair
81 to 100% of the threshold for age or usage
Poor
exceeds 100% of the threshold for age or usage or if
read/write errors have occurred on this medium
The thresholds used to calculate the status of a medium are described here:
Medium valid
for
Maximum
number of
overwrites
The age of a medium is calculated as the number of
months that have elapsed since you initialized the
medium. For this age to accurately represent the age of
your media, we recommend that you use new media when
creating a media pool. The default threshold is 36 months.
Once a medium is older than 36 months, it is marked as
having a poor status.
The usage of the medium is defined as the number of
overwrites to the medium. For OmniBack II to accurately
calculate the condition of your media, we recommend
using new media when creating a media pool. The default
threshold is 250 overwrites. Once a medium has more
than 250 overwrites, it is marked as poor.
3-21
Configuring a Cell
Media Management
Cleaning Tape
With this magazine pool attribute you specify that the device in this pool has
cleaning tape configured in the last slot. This slot will not be used for
backup.
See “The concept of logical devices” on page 3-25 and “Magazine pool” on
page 3-15 for more information on the cleaning tape support.
Managing media compatibility
OmniBack II recognizes the following media types:
•
•
•
•
•
•
•
tar
cpio
OmniStorage
OmniBack and OmniBack/Turbo
HP-UX filesystem
OmniBack II
ANSI labeled tapes
NOTE:
Detection of media types on Solaris is not reliable due to the use of different block
sizes.
Do not rely on OmniBack II to recognize other media types.
You may want to set the InitOnLoosePolicy option in the global options file.
The medium will not be automatically initialized and data will not be overwritten in
case it is not recognized. See “OmniBack II Global Options File” on page 7-11 for
instructions.
NOTE:
If you try to read from a medium that was written using hardware compression with
a device that does not support compression, OmniBack II will not recognize the DAT
and cannot read its data, but will overwrite it with the next backup.
3-22
Configuring a Cell
Media Management
The initialization policy for these backup media formats is shown in the
table below.
Format
OmniBack II
options
Overwrite
unknown
irrelevant
yes
tar, cpio, OmniStorage, OmniBack I,
ANSI label, filesystem
no-force
no
tar, cpio, OmniStorage, OmniBack I,
ANSI label, filesystem
-force
yes
OmniBack II
no-force
no
OmniBack II
-force
dependent on the
protection set for
the backup
session
Table 3-1
OmniBack II Overwrite Policy
OmniBack II only writes to media initialized by OmniBack II. You cannot
force the overwrite of another backup media format during a backup. This
helps avoiding accidental overwrites of other backup media.
Media Identification
Adding a medium to a media pool consists of initializing the medium. When
a medium is initialized, OmniBack II assignes two identifiers to the medium:
media ID
A unique ID created by OmniBack II, 32 characters in
length.
media label
A user-defined unique label, up to 32 characters in length.
If the device supports barcode, the media label becomes a
3-23
Configuring a Cell
Media Management
barcode.
For identification on other systems, an ANSI X3.27 label is also stamped on
the tape.
These labels identify the medium as an OmniBack II medium. When loading
a medium for backup or restore, OmniBack II checks the medium for the
media ID. When it finds the media ID, it checks the media management
system to see what information it has about this medium. This information
tells OmniBack II whether the requested action is allowed for this medium.
For example, if you are trying to write a new backup to this medium, the
media management system checks whether the expiration date for the data
already contained on this medium has passed.
OmniBack II displays the media ID, the media label, and the location when
prompting you to mount a medium. OmniBack II knows the media ID of a
mounted medium and does not display a mount prompt if the requested
medium is already in the device.
3-24
Configuring a Cell
Configuring and Using Logical Devices
Configuring and Using Logical Devices
OmniBack II introduces the concept of logical devices that represent
physical devices with various user defined characteristics such as location or
media mount handling. A logical device is associated with a media pool
used by OmniBack II to back up and restore data. Since backups and
restores are done using logical devices, you must configure your logical
devices before you can do a backup or restore.
For more information on configuring and using exchangers, see Chapter 9,
“Using Complex Devices and Exchangers,” on page 9-1.
The concept of logical devices
A logical device is a group of one or more physical devices of the same type
that are treated as one device by OmniBack II. Backups and restores are
done using logical devices.
For standalone devices, a logical device may consists of linked physical
devices that allow you to do unattended backups. When the medium in a
physical device is full, the backup automatically continues using the next
physical device in the logical device.
Note the following requirements for logical devices:
•
All the physical devices in a logical device must be of the same type (for
example, DAT)
•
All the physical devices in a logical device must exist on the same system
•
One physical device can only belong to ONE logical device
OmniBack II allows you to link similar physical devices and treat them as
one logical device. Note the following requirements for configuring and
using physical devices:
•
•
Each physical device can be a part of only one logical device.
One exchanger device (a device with repository slots) must be a separate
3-25
Configuring a Cell
Configuring and Using Logical Devices
logical device. You cannot create a logical device containing several
exchangers. For more information on using exchangers, see Chapter 9,
“Using Complex Devices and Exchangers,” on page 9-1.
Sometimes, the logical device for an unattended backup is unavailable (for
example, the logical device cannot connect to the host). If this logical device
has an auxiliary device configured, the backup will use the auxiliary logical
device. This auxiliary device can exist on any node in the cell.
Note the following requirements for configuring and using auxiliary devices:
•
They must be of the same device type as logical devices.
•
They cannot be used for backup at the same time as the primary devices.
The auxiliary device must be idle and ready for backup in case the primary
device fails.
•
Primary devices must not be configured as auxiliary devices as this would
create an endless loop.
•
An auxiliary device can only be used if the Media Agent for the primary
device cannot be contacted. This means, that for example, the system
where the primary device is connected is not accessible or OmniBack II
is not installed properly. If the primary device is just switched off but the
Media Agent can be contacted, the auxiliary device will not be used.
For an up-to-date list of supported physical devices, see the HP OpenView
OmniBack II Software Release Notes.
Cleaning tape
HP OmniBack II supports HP SureStore 12000e exchangers with cleaning
tapes.
For HP SureStore 12000e devices, OmniBack II can detect a cleaning tape.
If a cleaning tape is loaded during backup or restore, OmniBack II
recognizes the cleaning tape. When during backup the device reports that the
head is dirty, OmniBack II checks the exchanger’s repository for the
cleaning tape. If it is there, the device is cleaned automatically. If there is no
cleaning tape, the user is prompted to clean the device. Note that this option
3-26
Configuring a Cell
Configuring and Using Logical Devices
is by default turned off and no automatic tape cleaning is done. You must set
the OB2CLEANME=1 variable in the .omnirc file on the system where
the Media Agent for this device is running.
For other devices and HP SureStore 12000e without a cleaning tape, the
backup is paused and you are prompted to clean the drive. Once you have
cleaned the drive, you can resume your backup.
You cannot continue your backup without cleaning the drive, since the
backup may fail because data may not be correctly written and stored on the
media.
WARNING!
Cleaning tapes on SUN systems are not supported by OmniBack II.
If OmniBack II detects and inserts the cleaning tape, the tape driver enters
an undefined state and it may be necessary to reboot the system. You must
load the cleaning tape manually when requested to.
Configuring logical devices
HP OmniBack II stores information about each logical device in the
database.
Each logical device has the following characteristics:
•
•
•
•
•
•
•
•
•
Logical device name
Media class
Media pool
Host name
Auxiliary device
Logical device description
Unix Filename for each physical device in this logical device
Concurrency level
Mount Prompt Script
Logical devices are configured in the Logical Device
Administration window. From the OmniBack II Main window choose
Devices.
3-27
Configuring a Cell
Configuring and Using Logical Devices
Graphical User Interface Method
Logical devices are configured using the Logical Device Editor
window.
1. At the OmniBack II Main Menu click on Devices to open the HP
OmniBack II - Logical Device Administration window. This
window shows the logical devices that are currently configured for this
cell.
2. Select Edit → Create Logical Device and select your device type from
the cascading menu. The Logical Device Editor window for the
selected device type appears.
3. Enter required information. Use the ... button to get a list of options where
available. See online help for a detailed description of fields.
If you are configuring OmniBack II for the first time, the only media pool
shown in the list of media pools is the default pool for this device type (for
example "default DAT"). To configure a logical device, you have two
choices:
NOTE:
For SCSI II Exchangers you must first create an appropriate pool for the exchanger
and then configure a separate logical device for each drive in the exchanger.
•
•
You can create an empty pool (a pool without media), and use this pool
when configuring the logical device. You can then add media to this pool
after you have created the logical device. This is the recommended
method. See “The Media Pool approach” on page 3-13 and online help for
instructions.
You can use the default pool in order to complete the configuration of this
logical device. After you have configured the logical device, you can
configure an additional pool and assign a device to it. See “The Media
Pool approach” on page 3-13 and online help for instructions how to
create the media pool. Use Edit → Modify… to assign a device to the
pool.
We recommend that each pool is used by one logical device only.
3-28
Configuring a Cell
Configuring and Using Logical Devices
Text Editor Method
The command-line method for creating a logical device consists of two
steps:
NOTE:
•
creating the text file for the logical device
•
uploading the text file information into the database
If you are modifying a logical device using the command-line method, you must first
retrieve the database record for this logical device by using the omnidownload
command. For more information see the manpage for this command.
The two steps for creating a logical device using the command-line method
are described here:
1. Create the logical device file. OmniBack II provides you with template
files that you can edit and save. However, you can also create your own
logical device files, using the same format as the template files. This step
describes how to edit and save the template files.
Edit and save the template file using an ASCII text editor. Sample
template files for each type of supported logical device are located in the
directory /usr/omni/config/devices/ (on HP-UX 9.0x) or
/etc/opt/omni/devices (on HP-UX 10.x). The template files are
listed here:
tpexchanger
tpextcont
tpjukebox
tpstack
tpstalone
SCSI II exchanger
external control
magneto-optical jukebox
stacker
standalone
Each logical device template includes a variable labeled TYPE. This
variable indicates the device type.
0
1
2
3
5
6
DAT standalone
DAT SCSI II exchanger
DAT stacker
Magneto-Optical standalone
Magneto-Optical stacker
Magneto-Optical Pre 2.0 SCSI II exchanger
3-29
Configuring a Cell
Configuring and Using Logical Devices
10
11
12
13
14
15
16
17
18
19
20
22
23
24
25
26
27
28
29
31
32
33
Exabyte standalone
Exabyte stacker
Exabyte Pre 2.0 SCSII II exchanger
3480 standalone
3480 stacker
3480 SCSI II Pre 2.0 exchanger
File standalone
Reel tape
File exchanger - device files
DAT script exchanger
MO script exchanger
Exabyte script exchanger
3480 script exchanger
DLT standalone
DLT Pre 2.0 SCSI II exchanger
DLT stacker
DLT script exchanger
DAT SCSI II exchanger
MO SCSII II exchanger
Exabyte SCSI II exchanger
3480 SCSI II exchanger
DLT SCSI II exchanger
2. Upload the logical device file you created into the database using the
omniupload command. For more information see the manpage for this
command.
Configuring Typical Devices
This section provides examples how to configure some typical devices. Note
that OmniBack II offers great flexibility in combining logical devices and
the media pools concept. Make sure to read all relevant sections in order to
make a full use of this flexibility.
The following instructions are provided as guidelines only.
Standalone devices
Standalone devices are, for example, simple one drive DAT or ExaByte
devices.
1. Configure a media pool for standalone devices - a standard pool. See
“Standard pool” on page 3-14 and “Creating a media pool” on page 3-38
for more information. OmniBack II provides a default standalone pool
you can use instead of creating your own to simplify the configuration.
3-30
Configuring a Cell
Configuring and Using Logical Devices
You can later move media from the default pool to your own pool.
2. From the Logical Device Administration window select Edit →
Create Logical Device → Standalone to open the Logical Device
Editor - Standalone window.
Figure 3-1
Configuring a Standalone Device
3. Enter the required information. Use the ... button to get a list of options
where available. See online help for a detailed description of fields.
If you are configuring OmniBack II for the first time, the only media pool
shown in the list of media pools is the default pool, for example, default
DAT.
We recommend that each pool is used by only one logical device.
3-31
Configuring a Cell
Configuring and Using Logical Devices
Magazine devices (HP SureStore 12000e)
Magazine devices are devices that treat all media as a single unit, a
magazine, and support operations on this magazine, for example, the HP
SureStore 12000e.
1. Configure a media pool for magazine devices - a magazine pool. See
“Magazine pool” on page 3-15 and “Creating a media pool” on page 3-38
for more information. OmniBack II provides a default magazine pool you
can use instead of creating your own to simplify the configuration. You
can later move media from the default pool to your own pool.
You may also configure automatic head cleaning with provided cleaning
tape support.
2. From the Logical Device Administration window select Edit →
Create Logical Device → Magazine & Pre 2.0 SCSI II Exchanger to
open the Logical Device Editor - Pre 2.0 SCSI II Exchanger
window.
3. Enter the required information. Use the ... button to get a list of options
where available. See online help for a detailed description of fields.
If you are configuring OmniBack II for the first time, the only media pool
shown in the list of media pools is the default pool, for example, default
magazine.
We recommend that each pool is used by only one logical device.
SCSI II exchangers with repository control
OmniBack II introduces a new type of media pool to optimally manage and
control large exchangers. See “The Media Pool approach” on page 3-13 for
more information. In this configuration, each exchanger must have its own
exchanger pool that tracks media used in the exchanger.
See “Configuring a SCSI-II Exchanger” on page 9-7 for more information
on exchanger devices.
1. Configure a media pool, an exchanger pool. See “Exchanger pool” on
3-32
Configuring a Cell
Configuring and Using Logical Devices
page 3-17 and “Creating a media pool” on page 3-38 for more
information. Note that there is no default exchanger pool, as this pool
represents the repository and media used by a specific exchanger.
2. From the Logical Device Administration window select Edit →
Create Logical Device → SCSI II Exchanger to open the Logical
Device Editor - SCSI II Exchanger window.
Figure 3-2
Configuring a SCSI II Exchanger
3. Enter the required information. Use the ... button to get a list of options
where available. See online help for a detailed description of fields.
As a media pool choose the exchanger pool created in step 1.
3-33
Configuring a Cell
Configuring and Using Logical Devices
Pre 2.0 SCSI II exchangers
If you want to share the same media pool with several exchangers, you must
configure the exchangers as Pre 2.0 SCSI II exchangers and use a standard
pool. In this configuration you lose the possibility of tracking media in the
exchangers. See “The Media Pool approach” on page 3-13 for more
information on the new media pool concept.
See “Configuring a Pre 2.0 SCSI-II Exchanger” on page 9-13 for more
details on exchangers devices with a standard pool.
1. Configure a media pool, an standard pool. See “Standard pool” on
page 3-14 and “Creating a media pool” on page 3-38 for more
information.
2. From the Logical Device Administration window select Edit →
Create Logical Device → Magazine & Pre 2.0 SCSI II Exchanger to
open the Logical Device Editor - Magazine & Pre 2.0 SCSI
II Exchanger window.
3-34
Configuring a Cell
Configuring and Using Logical Devices
Figure 3-3
Configuring a Pre 2.0 SCSI II Exchanger
3. Enter the required information. Use the ... button to get a list of options
where available. See online help for a detailed description of fields.
As a media pool choose the standard pool created in step 1.
Rescanning a Device
For exchangers, rescan updates the database with correct information on the
data in the exchanger. Use this function when someone has manually
changed media in the exchanger. You can rescan either a single slot of an
exchanger or the complete device.
For standalone devices, rescan checks and displays information on what
media is in the device.
1. Select a logical device in the Logical Device Administration
3-35
Configuring a Cell
Configuring and Using Logical Devices
window and choose Edit → Rescan. The same functionality is available
from the Media Management window by choosing Edit → Scan.
The Rescan window allows you to select either a single slot or the
complete device. The Messages field of the window displays progress
messages during the rescan.
3-36
Configuring a Cell
Media Management Tasks
Media Management Tasks
Media management tasks are accessible using the Edit and View menus in
the Media Management window for the pool you select.
Migrating from a Previous Media Pool
With the A.02.00 release, OmniBack II introduces new types of media pools.
A new exchanger pool allows better management of exchanger devices, and
we therefore recommend migrating to this pool. See “The Media Pool
approach” on page 3-13 for more information.
Although you can still use your existing configuration, we recommend
migrating to the new exchanger pool.
Migrating to the new exchanger pool is a simple process:
1. Create a new exchanger pool. See “Creating a media pool” on page 3-38
for instructions.
2. Add exchanger’s slots you want to use to the newly created pool. See
“Adding slots to an exchanger pool” on page 3-38 for instructions.
3. Create a new logical device using this newly created pool. See
“Configuring logical devices” on page 3-27 for instructions.
4. Move media from the old pool to the newly created exchanger pool.
5. Scan the media in this pool to update the database information. See
“Scanning media” on page 3-45.
3-37
Configuring a Cell
Media Management Tasks
Creating a media pool
When adding media to pools, we recommend using new (unused) media.
This is because OmniBack II records the number of reads/writes to media
and uses this information to calculate the age and condition of the media. If
you add used media to a pool this information will be incorrect.
Figure 3-4
The OmniBack II Media Pools Window
1. From the Media Pools window select Edit Create Pool… and
choose the appropriate pool. The Media Pools - Pool Editor
window appears.
2. Enter the required information. See online help or “Pool attributes” on
page 3-19 for more information on specific fields.
Usually, you now add media to the pool. See below for instructions.
Adding slots to an exchanger pool
After you have created an exchanger pool, you must assign the slots of the
exchanger to this pool. You may add all slots or just a range. Note that the
slots used by OmniBack II should not be used by any other application.
3-38
Configuring a Cell
Media Management Tasks
OmniBack II will not overwrite existing data on the medium in the slot (see
“Managing media compatibility” on page 3-22 for more information), but
we cannot guarrantee that other applications will not overwrite OmniBack II
data on the media.
1. From the Media Management window choose Edit → Add slots...
2. Specify a range of slots in the exchanger to be used by OmniBack II.
Slots are displayed in the Media Management window, but the status of the
media is unknown until the media is used for the first time.
Adding media to a pool
This task is performed on exchanger pools only. Once you have created a
pool, you can add media to the pool. It is recommended that you add
sufficient media to the pool to allow unattended backups.
NOTE:
Once you have initialized a medium, you can only change the Medium Label and
Location in the database. To change the label and location on the medium itself,
you must reinitialize the medium.
1. Double-click on the pool name in the list of media pools to open the
Media Management window.
2. If you are initializing media in a magazine pool, you can choose between
initializing the whole magazine or a specific medium. Choose Edit → Init
Medium or Edit → Init Magazine.
3. Enter the required information in the fields. See online help for more
information.
3-39
Configuring a Cell
Media Management Tasks
Verifying a medium
Verifying a medium checks whether the data on the medium is valid.
OmniBack II does this by performing the following:
•
•
Checks the OmniBack II tape format headers
If the CRC [Cycle Redundancy Check] option is used, it recalculates the
CRC and compares it to the one stored on the medium.
To verify a medium, follow these steps:
1. In the Media Management window select a medium.
2. Choose Edit → Verify and select an appropriate logical device.
If verification was successful, the medium is in a good condition and the
counter used to define the media condition is reset to zero.
Erasing a medium
The Erase Medium option is only available if you have selected a pool
containing magneto-optical media. The action is used to pre-erase a
magneto-optical platter, which significantly increases the speed of the
backup because it eliminates the need for erasing the platter during backup.
To pre-erase a medium, follow these steps:
1. In the Media Management window select a medium/slot.
2. Select Edit → Erase and confirm your decision.
3-40
Configuring a Cell
Media Management Tasks
Modifying a medium
Modifying a medium lets you change the label or location of the medium in
the OmniBack II database. The label is an identifier that you have assigned
to the medium. Changing the location is useful if you send media to offsite
storage.
Modifying a medium only changes the information about this medium in the
OmniBack II database, but the medium itself remains unchanged.
To modify a medium, follow these steps:
1. In the Media Management window select a medium/slot.
2. Choose Edit → Modify from the menu bar to open the Pool Editor Modify Medium window.
3. Enter the new information in the Label and Location fields.
Viewing information about a medium
This action lets you view information about the usage and condition of a
medium. The condition of the medium affects your ability to write to a
medium and read the data contained on the medium. This helps you
determine when a medium should be replaced.
The Pool Editor - Medium Info window provides the following
information for the medium:
•
•
•
•
•
•
•
•
•
Medium Label
Medium ID
Medium location
Number of writes
Number of overwrites
Total Space (specified)
Used Space (calculated)
Medium initialization date and time
Last access date and time
3-41
Configuring a Cell
Media Management Tasks
•
•
•
Last write date and time
Last overwrite date and time
Detailed information for the objects on the medium
1. In the Media Management window select a medium.
2. Choose Edit → Info to open the Pool Editor - Medium Info
window.
Moving a medium
OmniBack II lets you move a medium from one pool to another pool. This is
useful if you want to use the medium on another device. Note that the new
pool must use the same type of media, for example DAT tapes.
1. In the Media Management window select a medium/media.
2. Choose Edit → Move to open the Pool Editor - Move Medium
window.
3. Click on a pool in the list to select the name of the destination pool. This
pool name appears in the field labeled Move Medium to this Pool:.
4. Click on the OK button to confirm the move and return to the Media
Management window. Note that the medium that was moved no longer
appears in the list of media for this pool.
Recycling media/Removing protection
OmniBack II keeps track of data on every used medium. When configuring
backup, you can protect your data from overwriting by newer backups for a
specific time. Recycling removes the data protection from backed up data,
thus allowing OmniBack II to overwrite it during the next backup. The data
on the media remains unchanged until the next backup.
To recycle a medium, follow these steps:
1. In the Media Management window select a medium/media.
3-42
Configuring a Cell
Media Management Tasks
2. Select Edit → Recycle and confirm your decision.
Exporting a medium
Exporting a medium removes the information about a medium and its
contents from the database. OmniBack II no longer knows that this medium
exists. The medium and the data it contains remain unchanged.
You need to export media when you want to move it to another cell. You
export it on one cell and import on another.
Media that hold protected data cannot be exported. You must recycle the
media first.
NOTE:
If you plan to export media, you should export all the media of a backup session. If
a backup session spans several media and you only export one medium of the set,
there will be inconsistencies in the OmniBack II database. These inconsistencies will
cause problems when you try to restore data from this backup session. These
problems occur because the backup session is in the database, but the data you want
to restore is on media that is no longer available. This generates an error message that
the session is not complete.
1. In the Media Management window select a medium.
2. Choose Edit → Export and confirm your action.
The medium that you exported no longer appears in the list of media for this
pool.
3-43
Configuring a Cell
Media Management Tasks
Importing a medium
The import of media rereads all data about the sessions that are on the
medium to the OmniBack II database.
If a session spans more than one tape, all tapes for this session should be
imported to the database. If you do not import all the tapes of the session, the
session will be only partially imported. This could cause problems at restore
time. If you select to restore from a partially imported session, you will get
the following message:
Full restore chain is broken or media needed for restore missing
You need to import media when you want to move media from one cell to
another. You export media on one cell and import on another.
Note that during import OmniBack II reconstructs the detail catalog of a
session. It read the label, location, and protection information from the
medium to the database. If this information changed between the
initialization and export of the medium, this information is lost, as
modifications are only done to the database and not to the medium.
Additional attributes such as object or media size will not be reconstructed
during the import.
To import a medium, follow these steps:
1. From the Media Management window choose Edit → Import… to
open the Media Management - Import window.
For exchangers, you must select slots first.
2. Enter the name of the logical device where the medium is loaded. To
select the logical device from a list, click on the ... button to the right of
the field.
3. Specify the amount of information to be logged. See “Description of
backup options” on page 4-22 for more information.
4. Click on the Import button to begin the import process. During the import
process, messages appear in the Messages field.
3-44
Configuring a Cell
Media Management Tasks
When the import is complete, you return to the Media Management
window. The imported medium now appears in the list of media for this
pool.
Scanning media
Scanning media checks the format of media inserted in alogical device.
1. From the Media Management window choose Edit → Scan...
For exchangers, you must select slots first.
2. Select the logical device where the media resides.
The Quick scan using barcodes option (default OFF) should be
used with exchangers supporting barcodes. OmniBack II does not read
each media label but only checks barcodes of media in the exchanger
thus saving time.
When the scan process completes, the Media Management window is
updated with information about the scanned media.
Removing exchanger slots and media from the exchanger pool
This task applies to exchanger pools only. After you have added specific
slots to an exchanger pool, you may want to remove some slots and use them
with some other application. Some slots may also contain non-OmniBack II
media you do not want to see in your exchanger pool.
1. In the Media Management window select the slot you want to remove
from the pool.
2. Choose Edit → Remove and confirm your decision.
The removed slot disappears from the list in the pool.
3-45
Configuring a Cell
Media Management Tasks
Entering media in an exchanger
This task applies to exchanger pools only. We recommend using this
procedure to enter new media in an exchanger. This allows OmniBack II to
maintain full control over the status of media in the exchanger.
In order for OmniBack II to fully recognize and control each non-OmniBack
II medium, the medium must have barcodes. If entered medium has a
barcode, OmniBack II searches in the database for the medium with the
same barcode and updates the repository information in the database
accordingly.
1. Enter a medium in the slot used to enter or eject media from the
exchanger.
2. In the Media Management window select a slot with Unknown or Empty
status where you want to enter the medium and choose Edit → Enter.
Ejecting media from an exchanger
This task applies to exchanger pools only. We recommend using this
procedure to eject media from an exchanger. This allows OmniBack II to
maintain full control over the status of media in the exchanger.
In order for OmniBack II to fully recognize and control each non-OmniBack
II medium, the medium must have barcodes.
1. In the Media Management window select the slot from which you want
to eject media.
2. Choose Edit → Eject to open the Eject Slots window.
Specify a Logical Device and a New Location. The new location
can be any text that can help you to find the medium later and will be
displayed in the Media Management window.
The status changes to empty.
3-46
Configuring a Cell
Media Management Tasks
Viewing media scheduled for backup
OmniBack II provides you with a way to check which media are currently
scheduled for backup. Using the View selection, you can choose between a
number of different views.
When you select one of these views, the list of media in the Media
Management window changes to the view you selected.
Additionally, you may want to sort your media using the Sort selection of
the Media Management window.
3-47
Configuring a Cell
Configuring User Classes and Users
Configuring User Classes and Users
How you configure your user classes and users has a major effect on the
security of a cell. Access to backup and restore functions gives a user access
to the data stored in that cell. Backing up data and restoring it to a different
disk is essentially the same as copying the data from one disk to another.
Data that can be backed up can also be removed and reloaded on a different
system. Therefore, it is recommended that you carefully plan what backup
and restore functions you give to your OmniBack II users.
Each OmniBack II user is a member of a user class and has the access
permissions of this user class. A user class is a group of users with specific
permissions to perform backup and restore tasks.
The following default classes are provided:
User
Class
Access Rights
admin
allowed to configure OmniBack II and do backup, restore
and all other available operations. After installation only the
root user on the Cell Server is configured as the
administrator.
senioroperator
allowed to do all operations except configure OmniBack II
operator
allowed to start backups and respond to mount requests
end-user
allowed to do restores of his/her own objects. In addition,
users can monitor and respond to mount requests for their
own restore sessions.
Table 3-2 OmniBack II Default User Classes and their Access Rights
3-48
Configuring a Cell
Configuring User Classes and Users
After the initial installation, all default user classes are empty except for the
admin class. The Unix root user on the Cell Server is automatically
configured as the OmniBack II administrator. All other categories can
remain empty if you want to restrict access to OmniBack II. You can add as
many additional user classes as you need for your environment.
Users are defined by their Unix login name and group. Wildcards (*) may be
used for hostname, user name and the group ID. There is no pattern
matching - only the asterisk (*) is a valid wildcard.
NOTE:
Admin capabilities are very powerful and they basically mean that the administrator
has root capabilities on the whole cell. Normally, only the root user on the Cell
Server should have Admin capabilities
You can use HP OmniBack II without modifying the default user
configuration. However, for security reasons it is recommended that you
verify (and if necessary modify) the default user configuration.
Configuring user classes
OmniBack II user classes are configured using the
User Administration - Add Class window. To access this window,
follow these steps:
1. At the OmniBack II Main menu, click on Users to open the
HP OmniBack II - Users Administration window.
2. Select Edit → Add → Class from the menu bar. The
User Administration - Add Class window appears.
3-49
Configuring a Cell
Configuring User Classes and Users
Figure 3-5
OmniBack II Users
Creating a user class consists of the following steps:
1. Defining a user class name
2. Defining a description
3. Selecting the access rights
4. Confirming the configuration of this user class
1. Enter a user class name in the Class Name field. The user class name
must be alphanumeric and can have a maximum of 16 characters without
spaces.
2. Enter a description in the Description field. The description field can
contain any characters, is a maximum of 64 characters long, and spaces
are allowed.
3. Select the access rights by clicking on the button to the left of the text. You
can scroll up and down to view the complete list of access rights. See
below for a description of the access rights.
4. Click OK to confirm this configuration.
Repeat these steps to add additional user classes. For specific information on
adding, modifying, and deleting user classes, see the online help system.
3-50
Configuring a Cell
Configuring User Classes and Users
The access rights are described below:
Start Backup
Start datalist
backup
Save datalist
allows a user to perform a backup of only their own data.
Note that this access right automatically also allows the
user to monitor and abort his/her own session.
allows a user to perform a backup using a datalist. This
access right allows the user to start a backup using any
datalist and therefore back up the data objects listed in
any datalist.
allows a user to create, schedule, modify, and save own
datalists.
Backup as root allows user to back up any object using the root
(superuser) login. Note that this is a powerful access right
that can affect the security of your system!
Mount Prompt allows a user to respond to mount prompts for any active
session in the cell.
Start restore
allows user to restore only his/her own objects (objects
for which he/she is the Unix owner). This access right
automatically also allows the user to monitor and abort
own session.
Restore to other
hosts
allows a user to restore objects to a host other than the
host where the object was backed up. The user must have
login permission on the destination host. Note that this is
a powerful access right.
Restore from
other users
allows the user to restore files belonging to another user.
Note that this is a powerful access right.
Restore as root allows user to restore objects using root capabilities.
Note that this is a powerful access right.
Monitor
allows a user to view session information for all sessions
in the cell.
Abort
allows a user to abort any active session in the cell. Note
3-51
Configuring a Cell
Configuring User Classes and Users
that this is a powerful access right.
Install
User
configuration
Device
configuration
Media
configuration
Switch session
ownership
NOTE:
allows the user to add or delete OmniBack II modules
from the cell. OmniBack II recognizes only those hosts
which have OmniBack II modules installed. Note that this
is a powerful access right.
allows a user to add, delete, and modify user classes and
users. Note that this is a powerful access right.
allows a user to create, delete, modify, and rename logical
devices. This includes the ability to add a mount prompt
script to the logical device. Note that this is a powerful
access right.
allows a user to manage media pools and the media in the
pools. Note that this is a powerful access right.
(Note, this access right is only appropriate if the "Start
Datalist Backup" access right is also enabled. The access
right "Switch Session Ownership" allows this user to start
a backup using a datalist without becoming the owner of
the backup session, if an owner has been set for the
datalist. You set the owner of the datalist using the File →
Change Ownership selection in the OmniBack II Backup window. If this access right if OFF, the user who
starts the backup becomes the owner of the backup
session.
The Edit → Class → Modify menu selection can be used to both view and modify
the access rights of a user class. You cannot modify the Admin class.
3-52
Configuring a Cell
Configuring User Classes and Users
Configuring users
Each OmniBack II user is a member of a user class and has the access
permissions of this user class.
OmniBack II users are configured using the User Administration Add User window. To access this window, follow these steps:
1. At the OmniBack II main menu, click on Users to open the HP
OmniBack II - Users Administration window.
2. Click on User Class in the list to select the User Class for this user.
3. Select Edit → Add → User from the menu bar to open the User
Administration - Add User window.
To add a user, follow these steps:
1. Enter the Unix login (user name) in the Unix Login field. By clicking
the ... button you will get a list of users configured in the local
/etc/passwd file.
2. Enter the Unix group name in the Unix Group field.
3. Enter the real name (actual name) of the user in the Real Name field.
4. Enter the host name of the user in the Host Name field.
5. Confirm the user class name in the Class field. If you want to use an
alternative User Class, click on the field to activate it, and enter the
alternative User Class name. Note that you can only enter the name of an
existing User Class.
6. Click the Add button to confirm the new user.
Repeat these steps to add additional users.
3-53
Configuring a Cell
Configuring User Classes and Users
3-54
4
Backing Up Your System
4-1
Backing Up Your System
In This Chapter
This chapter explains the procedures, commands and options for tasks done
during most backup sessions. It explains how to do the following:
•
back up filesystems and rawdisks
•
create and use datalists
•
configure the scheduler for scheduled backups
•
change backup ownership
Before you run a backup, you should have configured at least one host and a
logical device. If the logical device is resident on a separate host, configure
that host as well. See Chapter 3, “Configuring a Cell,” on page 3-1 for more
information.
OmniBack II provides several ways to back up data:
•
scheduled or interactive
•
filesystem or rawdisk
•
full or incremental
To run a backup you must specify what you want to backup. This is done
using a datalist. A datalist is a list of objects (filesystems or rawdisks) to be
backed up. To have the datalist backed up on a periodic basis, set up a
backup schedule using the Scheduler.
Non-routine backups are generally run interactively.
The following sections provide you an overview of the scheduled and
interactive backups.
4-2
Backing Up Your System
Understanding Datalists
Understanding Datalists
A datalist allows you to group and save backup objects, especially
convenient when these objects are backed up on a regular basis. Instead of
defining the objects every time you run a backup, use the settings saved in
the datalist.
HP OmniBack II datalists are text files of program-like statements that
specify the objects you want to back up. Objects are either filesystems or
rawdisks. A datalist also defines backup policies (in the form of options) to
be followed by OmniBack II when the backup is run. For more details on the
syntax of a datalist, see the manpage for omnidatalist.
Datalists are ASCII files located in the/usr/omni/config/datalists
directory (on HP-UX 9.0x) or /etc/opt/omni/datalists directory (on
HP-UX 10.x). It is possible to configure your own datalist using a text editor.
However, we recommend you use the GUI to design your datalist due to the
complexity of the datalist syntax. HP OmniBack II provides a GUI datalist
editor which has been designed to guide you through the process of
configuring a datalist.
Once your datalist is set up, you only need to change it if your system
configuration changes or if you want to change backup options. OmniBack
II automatically supplies default values for any options you omit from the
datalist.
When you are creating a datalist, the minimum information required is:
•
what to back up
•
where to backup (to which logical device)
You can create a datalist using two methods; choose the one that is simpler
for your case.
4-3
Backing Up Your System
Understanding Datalists
Generate
Datalist
Creates a datalist for the local host or all hosts in your
cell. This option automatically configures datalists for a
complete backup of all filesystems on the local host, of all
hosts in the cell of the OmniBack II database. You can
later modify the datalists to suit your specific
requirements.
Create Datalist You create the whole datalist from scratch, specifying all
options, or create a barlist for different integrations. For
more information on integrations see HP OpenView
OmniBack II Integration Guide.
Figure 4-1
OmniBack II Backup window
Generating Datalists
Creates a datalist for the local host or all hosts in your cell. This option
automatically configures datalists for a complete backup of all filesystems
on the local host, of all hosts in the cell, or of the OmniBack II database. You
can later modify the datalists to suit your specific requirements.
4-4
Backing Up Your System
Understanding Datalists
1. Choose Backup at the OmniBack II Main menu to open the OmniBack
II Backup window..
2. Select Edit → Generate Datalist… and choose whether you want to
create the datalist for a local host only, for all hosts in the cell, or for the
OmniBack II database. You only need to enter a name of the datalist and
a backup device. Your datalist will be added to the OmniBack II
Backup window.
3. If you need to modify this datalist see the online help or “Modifying a
datalist” on page 4-6 for instructions.
Creating Datalists
You create the datalist from scratch, specifying all options, or create barlists
for different database integrations (see HP OpenView OmniBack II
Integration Guide for more information.)
To create a datalist do the following:
1. Choose Backup at the OmniBack II Main menu to open the OmniBack
II Backup window..
2. Select Edit → Create → OmniBack II Datalist to open the OmniBack
II Backup Editor window.
3. Select Object → Add and choose from the list which type of backup you
want to do.
4. Enter the defining parameters for the object(s) you want to backup.
For detailed explanations on these parameters and their options, see the
online help and “Using Backup Options” on page 4-19.
4-5
Backing Up Your System
Understanding Datalists
Modifying a datalist
NOTE:
If a cell is changed, for example, a node is removed, you only need to load and save
the datalist with this node. Otherwise, OmniBack II will generate a warning message.
1. Choose Backup from the OmniBack II Main menu to open the
OmniBack II Backup window.
2. Highlight the datalist you want to modify.
3. Select Edit → Modify to open the OmniBack II Backup Editor
window.
You can add, remove, or modify objects specified in the datalist, and add
backup options to each object or to the datalist. See “Using Backup
Options” on page 4-19 for more information.
4-6
Backing Up Your System
Configuring a Filesystem Backup
Configuring a Filesystem Backup
A filesystem is an organization of files on a mass memory device. A
filesystem is backed up so that the filename, file attributes, and the file
contents are stored on the media. The set of files backed up from a filesystem
in a specific session is called a "fileset." The fileset backed up from one
filesystem changes from backup session to backup session depending on
user activity on this filesystem (creation of new files, deletion of old files,
etc.) and whether a full or incremental backup has been run in this session.
OmniBack II keeps track of all backed up filesystems and knows when and
to which media each filesystem has been backed up. OmniBack II uses this
information when you request a file or filesystem be restored. A disk
containing a filesystem is identified by its hostname and mountpoint.
The procedure for running a scheduled filesystem backup involves four main
tasks:
1. selecting the filesystem to be backed up
2. selecting the backup options
3. setting up the backup schedule
4. previewing the backup
Selecting afilesystem to backup
1. Choose Backup at the OmniBack II Main window. The OmniBack II
Backup window appears.
2. Select Edit → Create Datalist from the menu bar. The OmniBack II
Backup Editor appears.
3. Select Object → Add → Filesystem to open the OmniBack II
Filesystem Backup window.
4-7
Backing Up Your System
Configuring a Filesystem Backup
Figure 4-2
Adding a Filesystem
4. Enter the hostname in the Hostname field or click on the ... button to see
a list of hosts.
5. Enter the mountpoint (which is the root directory of your mounted
filesystem) in the Mount Point field or press the ... button to choose the
mountpoint from a list.
6. Enter a description of this filesystem backup in the Description field. It is
useful to enter the mountpoint when you back up more than one disk on
the same system. This makes locating files easier.
7. Choose Exclude to exclude specific files and directories from the backup.
Note that you must enter the full pathname of files. For example, you may
want to exclude the /tmp directory.
4-8
Backing Up Your System
Configuring a Filesystem Backup
NOTE:
Use the Exclude option to exclude the configuration and database
directory/usr/omni/config (on HP-UX 9.0x) or /var/opt/omni/db and
/etc/opt/omni (on HP-UX 10.x) from your Full Filesystem backups. Because
the database is in use during the backup a backup of the database will be incomplete
and you will not be able to restore the database from this backup.
We also recommend excluding the NFS mounted filesystems to avoid error reports.
8. Click on the Options button if you want to specify one of the following
options:
Only
Enter the names of files to be backed up. You can use
wildcards when selecting the filename. For example, you
may want to back up only *.doc files.
Skip
Excludes files matching the entered file name criterion.
You can use wildcards to specify the file name. For
example, you may want to skip all *.o files in the
selected fileset.
9. Click on the OK to go back to the OmniBack II Filesystem Backup
windows.
10. To specify additional filesets to back up, enter the full path name and
file/directory in the field to the left of the Add button, and click Add.
If you do not specify a particular file or directory, the complete filesystem
is configured for backup.
To remove a file from the Files and Directories field, select the file or
directory and click on Delete.
11. Enter the name of the logical device to which the fileset will be backed up.
Enter it in the Backup Device field, or click on the ... button to select from
a list.
After you have configured one filesystem, set the backup options. You may
set up the options yourself or use the default values provided.
4-9
Backing Up Your System
Configuring a Filesystem Backup
Selecting the filesystem backup options
1. In the OmniBack II Filesystem Backup window, click on the
Filesystem Backup Options button to display and set the backup options
for this object. These options override the session default options of the
datalist. See “Using Backup Options” on page 4-19 for more details.
Choose Default to set options back to default settings.
2. Click on the OK button to set options and return to the Filesystem
Backup window.
3. Click on the OK button in the Filesystem Backup window to close this
window and return to the OmniBack II Editor window.
4. You can now set the options for the device where the backup will be
performed. Select the device in the Devices field in the OmniBack II
Backup Editor window and select Options → For Selected Device…
to open the Devices Options window. See “Device Options” on
page 4-25 for more information.
5. Click on the OK button to set options.
Now you can save the datalist and run the backup interactively or schedule
the datalist. See “Starting a backup of an existing datalist” on page 4-14 or
“Setting Up Scheduled Backups” on page 4-28 for more information.
4-10
Backing Up Your System
Configuring a Rawdisk Backup
Configuring a Rawdisk Backup
A rawdisk backup is a high-speed backup where the system does not track
the logical structure of the information stored on the disk. OmniBack II
accesses the sections using the character device filename
/dev/rdsk/filename that you specify. The hostname you enter provides
the location of the sections.
Scenarios in which a rawdisk backup is appropriate include:
NOTE:
•
full system backup intended for disaster recovery (full system restore)
•
system backup before a major software update; for example, a new
version of the operating system
•
duplication of a filesystem on another disk, if a direct disk-to-disk
connection is not possible; the two disks must be identical
•
backup of a disk that contains databases
Rawdisk backups larger than 2 Gb are not supported.
We recommend that you unmount a rawdisk before backup and mount it back later.
You can use Pre- and Post- exec commands to do that.
Although a rawdisk backup does not distinguish between file system disks
and database sections, OmniBack II still allows you to restore individual
files from a rawdisk backup if the rawdisk holds an HP-UX HFS filesystem.
Rawdisk sections are backed up in the order you enter them in the Rawdisk
Sections list. OmniBack II provides for network-wide rawdisk backup;
however, for maximum performance rawdisk backup is best run locally.
The procedure for running a scheduled rawdisk backup involves the same
four main tasks you perform for a scheduled filesystem backup:
1. specify the rawdisk to be backed up
2. select the backup options
3. set up the back up schedule
4-11
Backing Up Your System
Configuring a Rawdisk Backup
4. preview the backup
Specifying the rawdisk sections for backup
1. Choose Backup at the OmniBack II Main window. The OmniBack II
Backup window appears.
2. Select Edit → Create → OmniBack II Datalist to open the OmniBack
II Backup Editor window.
3. Select Object → Add → Rawdisk to open the OmniBack II Rawdisk
Backup window.
Figure 4-3
Adding a Rawdisk
4. In the Hostname field enter the hostname where the rawdisk is located or
click on the ... button to see a list of hosts.
5. In the Description field enter descriptive text for this backup object.
4-12
Backing Up Your System
Configuring a Rawdisk Backup
Enter up to 32 characters including spaces. Note that you cannot leave this
field blank.
6. In the field beside the Add button, enter the disk sections that you want
backed up. Enter the section character device name. On HP-UX systems,
all character device files are listed in the /dev/rdsk directory.
7. Click on Add to add the section to the Rawdisk Sections list.
To delete a section from the list, highlight a section in the Rawdisk
Sections list and click on Delete.
8. Enter the name of the logical device to which the rawdisk will be backed
up in the Backup Device field, or click on the ... button to select from a
list.
Selecting the rawdisk backup options
1. Click on the Rawdisk Backup Options button to display and set the
backup options for this object. See “Using Backup Options” on page 4-19
for more information.
Choose Default to reset the options back to the default values.
2. Click on the OK button to set options and return to the Backup Editor
window.
3. Next, set the options for the device where the backup will be performed.
Select the device in the Devices field in the Backup Editor window.
Choose Options → For Selected Device to open the Devices Options
window.
See “Device Options” on page 4-25 for more information.
4. Click on the OK button to set the device options and return to the
OmniBack II Backup Editor window.
Now you can save the datalist and run the backup interactively or schedule
the datalist. See “Starting a backup of an existing datalist” on page 4-14 or
“Setting Up Scheduled Backups” on page 4-28 for more information.
4-13
Backing Up Your System
Running Interactive Backups
Running Interactive Backups
Running an interactive backup means that the backup is performed
immediately and interactively.
•
You can run an interactive backup of an existing datalist. See “Previewing
your backup” on page 4-18 and “Starting a backup of an existing datalist”
on page 4-14 for instructions.
•
You can create a temporary datalist for this backup only and run the
backup of this datalist. See “Creating a temporary datalist” on page 4-17
for instructions.
Starting a backup of an existing datalist
If you have already configured datalists and you want to run an interactive
backup of an existing datalist, use the following procedure:
1. From the OmniBack II Backup window select the datalist you want to
back up.
2. Choose Actions → Backup → Selected Datalist… to open the
OmniBack II Backup Editor window.
3. Choose Actions → Start Backup.
4-14
Backing Up Your System
Running Interactive Backups
Figure 4-4
Running a Backup Session
The message field displays progress messages.
The maximum buffer for messages in the Message field is limited with
the configurable parameter MAXGUIMSG. See “OmniBack II Global
Options File” on page 7-11 for more information. If the maximum buffer
is exceeded, messages stop appearing in the Message field and the error
message “Output disabled” appears, but the backup continues.
TIP:
By pressing the right mouse button in the message area you get a pop-up
menu where you can search for specific messages.
4-15
Backing Up Your System
Running Interactive Backups
Status messages
The Messages field uses colors and status descriptions to help you identity
the progress and status of the backup.
The status is displayed for each object and datalist (session) separately.
Status of backup objects
The status of backup object shows how each object in the datalist has been
backed up.
Completed
All files have been backed up successfully without errors,
there may be some warning messages. The field is green.
Completed/Errors Some files were not backed up successfully. The field is
yellow.
Aborted
The Disk Agent has been aborted for the following
reasons:
• The user has aborted backup.
• The backup of all files has failed.
• The Pre- or Post- exec script has failed.
Usually files have not been backed up successfully. The
field is red.
Failed
OmniBack II could not run backup. There may be
something wrong with the configuration. No backup
exists. The field is red.
Status of datalist (session) backup
The status of datalist (session) shows status of the whole datalist backup.
The status of each object in the datalist directly affects status of the datalist.
Completed
All backup objects are backed up successfully without
errors, there may be some warning messages. The field is
green.
Completed/Errors Some objects have been backed up with errors (status
4-16
Backing Up Your System
Running Interactive Backups
Completed/Errors) but all objects have been backed up.
The field is green.
Completed/Failures At least one backup object has not been successfully
backed up (with status Aborted or Failed.) The field is
yellow.
Aborted
The session has been aborted. A backup of some objects
may exist. The field is red.
Failed
A backup of all backup objects has failed, or the user has
aborted the backup (status of all backup object is Aborted
or Failed.) No backup exists. The field is red.
Creating a temporary datalist
1. Use Action → Backup → Interactive to open the OmniBack II
Backup Editor window.
2. Use Object → Add → … and select an object you want to add. Enter
required information in the window.
Use the ... button to get a list of possible options where available.
Use the Exclude option to exclude the configuration and database
directory/usr/omni/config (on HP-UX 9.0x) or
/var/opt/omni/db and /etc/opt/omni (on HP-UX 10.x) from
your Full Filesystem backups. Because the database is in use during the
backup a backup of the database will be incomplete and you will not be
able to restore the database from this backup.
We also recommend excluding the NFS mounted filesystem to avoid
error reports.
For an explanation of backup options available for a specific object, see
the online help or “Using Backup Options” on page 4-19.
Repeat this procedure to specify as many objects as needed. You can modify
backup options for each specific object later if desired.
If you plan to use this datalist again, use the File → Save to save it.
4-17
Backing Up Your System
Running Interactive Backups
Previewing your backup
You can preview a selected backup session to see if you have properly
configured options and if the correct files are being backed up and to get an
estimation of how much space will be used for data.
NOTE:
The Pre- and Post- exec commands set for an object are executed during the preview
of a backup. If these commands affect your operations, you may want to preview your
backup first and then add the Pre- and Post- exec commands.
1. In the OmniBack II Backup Editor for a selected datalist choose
Actions → Preview Backup. Choose whether you want to preview a full
or incremental backup of selected objects. If this is the first time objects
are backed up, the preview will show a full backup.
Starting your backup
1. In the OmniBack II Backup Editor for a selected datalist choose
Actions → Start Backup. Choose whether you want to run a full or
incremental backup. If this is the first time objects are backed up, a full
backup will be run automatically.
The message field displays progress messages.
The maximum buffer for messages in the Message field is limited with
the configurable parameter MAXGUIMSG. If the maximum buffer is
exceeded, messages stop appearing in the Message field and the error
message “Output disabled” appears. See “OmniBack II Global Options
File” on page 7-11 for instructions how to change this value.
TIP:
By pressing the right mouse button in the message area you get a pop-up
menu where you can search for specific messages.
The Messages field provides colors and status descriptions to help you
identity the progress and status of the session. See the online help or “Status
messages” on page 4-16 for a description.
4-18
Backing Up Your System
Using Backup Options
Using Backup Options
HP OmniBack II offers a set of comprehensive backup options allowing fine
tuning of a backup. All options have default values which are appropriate for
most cases.
Figure 4-5
Backup Options
4-19
Backing Up Your System
Using Backup Options
For a datalist
Pre- exec commands
Post- exec commands
These commands are executed before and after a datalist backup. You
could, for example, set up that an email is sent to users or the
administrator with a list of backed up files.
To set the Pre- and Post- exec commands for the datalist, choose Options
→ For Selected Datalist from the OmniBack II Backup Editor
window.
Session Defaults for Filesystem Objects
compress
encrypt
private/public
report level (warning, minor, major, critical)
display statistical information
lock files during backup
do not preserve time attributes
ignore hard links
pre- exec command
post- exec command
protection
log files/directories to database
Default backup options for filesystem objects in a datalist. These options
are applied each time a filesystem is being backed up. You could, for
example, set the Compress option for all filesystem objects in the
datalist.
To set these options select Options → Session Defaults → For
Filesystem Objects from the OmniBack II Backup Editor window.
Session Defaults for Rawdisk Objects
compress
encrypt
private/public
4-20
Backing Up Your System
Using Backup Options
report level (warning, minor, major, critical)
pre- exec command
post- exec command
protection
Default backup options for rawdisk objects in a datalist. These options
are applied each time a rawdisk is being backed up. You could, for
example, set the Compress option for all rawdisk objects in the datalist.
To set these options select Options → Session Defaults → For Rawdisk
Objects from the OmniBack II Backup Editor window.
For Selected Object
Backup options for each specific object in a datalist. These options
override the session default options and are applied each time the specific
object is backed up. You could specify, for example, that one object in a
datalist does not use the Compress option, because the system is too slow
and the object contains only a small amount of data.
See above for a list of options depending on the selected backup object.
To set these options select the object in the OmniBack II Backup
editor window and choose Options → For Selected Object….
For Selected Device…
CRC Check
Concurrency
Media Pool
Prealloc List
To set these options select the device in the OmniBack II Backup
editor window and choose Options → For Selected Device….
4-21
Backing Up Your System
Using Backup Options
Description of backup options
Compress
Data compression lets you write data to the media in a compressed format.
Compression reduces the number of media required for a backup and
improves backup performance. By default, the compression option is set to
OFF. When software compression is turned on, OmniBack II compresses the
data on the node where the Disk Agent is running before it sends it over the
network. This reduces data traffic over the network. Depending on the the
type of data, compression ranges from 30% to 70% and is based on the
Lempel-Ziv 4.3 compression algorithm which is completely compatible
with the standard Unix compress utility. OmniBack II provides you with
the API (Application Programming Interface) which is used by the Disk
Agent to interface with the compression module. This allows you to
substitute your own special-purpose compression module for even higher
rates of compression. OmniBack II also supports devices which allow for
hardware data compression.
NOTE:
Most modern backup devices provide built-in hardware compression that you can
configure when adding the device to the system. See your device’s manual for
instructions. In this case do not use the compress option of OmniBack II, since double
compression only decreases performance without giving better compression results.
Encrypt
Open Systems and public networking make data security in large enterprises
essential. OmniBack II lets you encrypt file and rawdisk data so that it
becomes unreadable. Data is encrypted before it is transferred over the
network and before it is written to media. By default, the encrypt option is
set to OFF. OmniBack II offers a simple built-in XOR algorithm
implemented in a shared "C" program library. Since OmniBack II provides
the API used by the Disk Agent to interface with the encryption module, you
can substitute your own internal encryption algorithms for greater security.
Do this by writing your own encryption module. Compile the encryption
module into a library and substitute it for the default OmniBack II library.
4-22
Backing Up Your System
Using Backup Options
You must generate your own encryption module and key as well as a method
of distributing the keys. See the section, "Customizing the Encryption
Algorithm" in the chapter "Using Advanced OmniBack II Features" for the
format and other details relating to encryption.
Public/Private
This option lets you set access rights for restoring the data you back up. If a
filesystem is backed up with the Private option, it can be restored only by
you or the system administrator.
By default, this option is set to Private. Setting the option to Public lets
anyone restore your data.
Report Level
This option defines the level of errors that will be reported during a backup
session. Setting a level means errors of this and higher levels are reported.
For example, by setting the Minor button ON, all errors graded as minor,
major, and critical are reported in the Messages field. By default the level
is set to Warning. Messages keyed as Normal always appear in the
Messages field. See “Projecting the growth of the database” on page 8-4 for
a description how this option influences the OmniBack II database.
Log Files to Database
OmniBack II provides three levels of the amount of details on files and
directories stored in the database.
Log All
This is the default option. All backed up files are logged
to the database. This complete information allows you to
search for backed up files and allows OmniBack II to
quickly position on the tape when restoring a specific file.
However, this information may take a lot of space if there
are many files.
Log Directories Details on directories only are stored in the database. This
disables the search feature during restore and you will
only be able to browse directories. However, OmniBack
II still performs fast positioning because a file is located
on the tape near the directory where it actually resides.
This option is suitable for filesystems with many autogenerated files, such as news and mail systems.
4-23
Backing Up Your System
Using Backup Options
No Log
No details on files or directories are logged in the
database. You will not be able to search and browse files
and directories. The restore will take longer, because
OmniBack II cannot fast position on the tape but will read
from the start of the backup.
See “Projecting the growth of the database” on page 8-4 for a description
how this option influences the OmniBack II database.
Lock Files During Backup
This option defines how files are handled when the files are being used
during a backup session. If set to ON, files are locked during the backup
session, preventing the files from being modified during the backup. The
default setting is OFF.
Do Not Preserve Time Attributes
Whenever a file is opened (read, locked, etc.) the access time of the file is
changed. During backup each file in the specified fileset is opened, locked,
and read. When this option is set to OFF (default), the access time attributes
retain as they were before backup (they are reset to their original values).
When you set this option to ON, the access time values are changed to a
moment of backup.
Ignore hard links
Causes OmniBack II to traverse the filesystem tree only once, thus
significantly speeding up the backup process. Use this option when there are
no hard links in your directory. When this option is set to ON, OmniBack II
cannot estimate the size of the backup or display the percentage of the
backup finished. The default is OFF.
Pre- and Post-Exec
The Pre-Exec option allows you to enter a command to be executed before
the backup of a datalist or each object is initiated.
The Post-Exec option allows you to enter a command to be executed after
the backup of a datalist or an object has completed. You can customize the
pre- and post-exec commands to your environment.
See “Using Pre- and Post- exec functions” on page 7-6 for more
information.
4-24
Backing Up Your System
Using Backup Options
Protection
This option enables you to set periods of protection for the data you back up.
This prevents the backup data from being overwritten for the specified
period. The default value is Permanent.
The backup protections are:
•
None-Provides no protection. Data can be overwritten if the medium it is
stored on is reused.
•
Until-Data on the medium cannot be overwritten until a specified date.
You enter the year, month and day. Protection for the data will stop at
noon of the entered day. For example, if you enter in the entry field
provided the date 96/8/22, the data will be protected until noon, August
22, 1996.
•
Days-Data on the medium cannot be overwritten for the number of days
specified. For example, if you set the protection for 2 days, and the current
backup starts at 20:00 (8:00 PM) on Wednesday, the data will be protected
until Friday at 20:00.
•
Weeks-Data on the medium cannot be overwritten for the number of
weeks specified. For example, if you set the protection for 1 week, and the
current backup starts at 10:00 on Wednesday, the data will be protected
until 10:00 the following Wednesday.
•
Permanent-the data is permanently protected from being overwritten.
Device Options
Set these options for the selected backup destination device.
CRC Check
Set this option ON to have OmniBack II calculate the CRC (cycle
redundancy check) when a backup is run. The CRC check is an enhanced
checksum function that lets you confirm whether or not data has been
written correctly to the medium. To verify media written with the CRC
option, use the OmniBack II command omniver in the command line
interface.
4-25
Backing Up Your System
Using Backup Options
Concurrency
Concurrency allows more than one Disk Agent (up to five) to write to one
backup device. This helps OmniBack II keep the device streaming when it
can accept data faster than a disk can send it (for DAT devices, this value is
2.) OmniBack II provides default values for all supported devices.
Media Pool
This option selects the media pool where the backup will be done. If not
defined, a default pool will be used.
Prealloc List
The Prealloc list is a subset of media out of the media pool that OmniBack II
will use for backup. This list also determines the order in which the media
will be used. This order is important if you have selected the Strict media
allocation policy when you configured the logical device. In this case,
OmniBack II expects the sequence of the media in the device to correspond
with that specified in the Prealloc list. If the media is not available in
this sequence, OmniBack II issues a mount request.
Note that this option is ignored with magazine pools.
Transferring backup ownership
OmniBack II lets you change the owner of a backup session. For example, as
system administrator you may create a datalist and back up certain data for
other employees. Changing the ownership of a backup session lets someone
start a backup session for objects not owned by him/her.
Changing owner works only if the datalist is saved. If the datalist is changed
but not saved the backup is treated as an interactive backup and the owner is
not changed. This could result in a different kind of backup than expected.
To switch ownership follow the steps below:
1. Open the main OmniBack II Backup window and select the datalist
whose owner you want to change.
2. Choose File → Change Ownership to open the Session Ownership
4-26
Backing Up Your System
Using Backup Options
window.
3. In the Session Ownership window enter the user name, group and host
name of the user you want to transfer ownership to. No spaces are allowed
between characters in each entry.
The default ownership of the backup session and objects remains with the
original initiator of the backup session. See “Configuring users” on
page 3-53 for more detals.
4-27
Backing Up Your System
Setting Up Scheduled Backups
Setting Up Scheduled Backups
By setting up a schedule for a datalist, you can automate your backups to run
periodically at times of low system usage. With the scheduler, the backups
you set up will run on the days and at the times you specify without your
attendance as long as the devices and media are properly set. You can also
specify the dates (holidays) to be skipped. See “Skipping backups during
holidays” on page 4-34 for instructions.
When a backup session is started, OmniBack II tries to locate all the
resources needed (for example, the correct number of licenses, the devices,
and access rights to the database). If any of these resources are unavailable,
the session is marked as "queued". OmniBack II then tries again to get the
resource once a minute until the timeout is reached. When the resources
become available, one of queued sessions is randomly selected and started.
(Note that the sessions are not started in sequential order.)
If OmniBack II cannot get the necessary resources within the allotted time,
the session fails. The timeout is controlled by two parameters in the Global
file:
The SmWaitForDevice variable controls the timeout for the device and
the licenses. If the requested devices and necessary licenses are not
available before this time (in minutes) has elapsed, the session fails. The
default value is 60 minutes.
The SmWaitForDB variable controls the timeout for the OmniBack II
database. If the OmniBack II database is not available before this time (in
minutes) has elapsed, the session fails. The default value is 60 minutes.
See “OmniBack II Global Options File” on page 7-11 for more information.
You can set up a schedule using the GUI.
As soon as you have created and saved the schedule for a configured datalist,
the schedule is activated and OmniBack II will check the schedule. No
additional activation step is required.
4-28
Backing Up Your System
Setting Up Scheduled Backups
As the backup is running, the OmniBack II Backup Monitor window
may issue mount prompts requesting media be mounted. If no one will be
present to respond to mount prompts at times chosen to run a backup, set up
mount prompt scripts for automatic mount prompt handling - see the
“Automating Mount Prompt Handling” on page 7-3 for more information.
NOTE:
When more than 5 sessions are scheduled at the same time, the exceeding sessions
are queued. You can change this number by modifying the MaxBSessions variable
in the global options file.
Setting up backup schedules
There are two general methods of setting the start dates and times for the
backup:
•
Method 1: By specifying the exact start date(s) and time(s)
•
Method 2: By specifying regular intervals.
In Method 2, OmniBack II offers you two ways to specify the interval of
backups. Using:
•
predefined intervals
•
custom intervals
In all cases, you can schedule backups for up to a year in advance.
Make sure the cell you are setting the schedule for is within one time zone. If
the cell is spread over more than one time zone, OmniBack II will use the
time from the Cell Server.
4-29
Backing Up Your System
Setting Up Scheduled Backups
Setting up a backup for a specified date
Figure 4-6
Backup Scheduler for Setting Specific Dates
The following section explains how to do a backup on a specific date.
1. In the OmniBack II Backup window, highlight a datalist to be
scheduled.
2. Select Edit → Change Schedule from the menu bar. The Scheduler
window opens.
3. In the Scheduler window select the View pulldown menu, choose a
month (January-December).
4. Click on the date you want to run the backup on in the Scheduler
window to open the Daily Edit window.
5. Choose a Backup Type (Full or Incremental).
6. Select the starting time for your backup session. You can run a backup as
frequently as every fifteen minutes. OmniBack II issues error messages if
you select conflicting backup dates or times.
4-30
Backing Up Your System
Setting Up Scheduled Backups
The type of backup is indicated by the colors defined in the Legend at the
bottom of the Scheduler window.
Figure 4-7
Backup Scheduler for Setting Specific Times
For example, to set up an incremental backup on October 4th, at 3:30, follow
these steps:
1. Choose October from View menu.
2. Select day 4.
3. Select Backup Type Incremental.
4. Click on start time 3:30.
5. Click on the OK button. The main scheduler window will reappear with
the date you set up highlighted in blue, the Legend color for an
incremental backup.
6. Select File → Save from the menu bar.
Selecting a predefined periodic schedule
You can set up a schedule based on the frequency of the backup by choosing
a predefined or custom schedule (see the next section). You can use the
predefined schedule as templates to create your own customized schedules.
1. At the OmniBack II Main backup window, highlight a datalist to be
4-31
Backing Up Your System
Setting Up Scheduled Backups
scheduled.
2. Select Edit → Change Schedule from the menu bar. The Scheduler
window opens.
3. In the Scheduler window select Predefined from the pull down menu.
4. Choose one of the predefined schedules. They are as follows:
Daily intensive
OmniBack II runs a full backup at midnight, and two additional incremental
backups at 12:00 and 18:00 every day. This type of backup is intended for
database transaction servers and similar environments with intensive backup
requirements.
Daily full
OmniBack II runs a full backup every day at 21:00. This is intended for
backups of single workstations or servers.
Weekly full
OmniBack II runs a full backup every Friday and Incr-1 backups everyday at
21:00 from Monday to Friday. This is intended for small workgroup
environments.
Fortnightly full
OmniBack II runs a full backup every second Friday. Between these
backups, OmniBack II runs Incr-1 backups every Monday to Thursday, all at
21:00.
Monthly full
OmniBack II runs a full backup on the 1st of every month, Incr-1 backup
every week, and an incremental every other day. This is intended for
relatively static environments.
NOTE:
A time period is counted from January 1st 1970.
4-32
Backing Up Your System
Setting Up Scheduled Backups
Setting up a custom periodic schedule
1. At the HP OmniBack II main backup window, highlight the datalist to be
scheduled.
2. Select Edit → Change Schedule from the menu bar. The Scheduler
window opens.
3. In the Scheduler window select Edit → Periodic from the pull down
menu. The Scheduler-Periodic Tagging window opens.
4. In the Scheduler-Periodic Tagging window, select a Backup Type (Full
or Incremental 0-9).
5. Choose a Time Cycle (every, every 2nd, etc.)
6. Choose a Time Unit (day, month,etc.).
7. Enter the Backup Time using the 24-hour clock system.
8. Click Apply to apply the schedule. If there are scheduling conflicts,
OmniBack II will notify you with a message, so you can modify the
schedule.
Editing the daily or monthly backup schedule
Clearing a Schedule
To eliminate a schedule that you have already set up, select the Edit →
Clear → All in the Scheduler window.
Note that when you clear a schedule, you clear all the schedule settings of a
specified mode (full, incremental, or all) for the current year.
Undoing the Clear
To undo the clearing a schedule, in the Scheduler window, click Edit →
Undo.
4-33
Backing Up Your System
Setting Up Scheduled Backups
Changing the Start Date
To change the start date, simply follow the procedure for setting up a backup
for a specific date. See the section in this chapter, "Setting Up a Backup for a
Specified Date."
Skipping backups during holidays
HP OmniBack II uses the /usr/omni/config/Holidays file (for HP-UX
9.x) or /etc/opt/omni/Holidays file (on HP-UX 10.x) to enable you to
set up a holiday backup schedule. Therefore the file must be created before
you can use this feature. This file uses the same syntax as the standard UNIX
holidays file.
To skip running a backup during holidays, click on the Holiday button in the
Scheduler window. This instructs the scheduler to exclude from the
current schedule all dates found in the Holidays file. For example, if the
date January 1 is registered as a holiday, OmniBack II will not back up on
that date. If you have scheduled a full backup for January 1st, and an
incremental for January 2nd, OmniBack II will skip running the full backup
on January 1st but will run the incremental backup scheduled for January
2nd.
Overriding scheduled backup options
Use the Override option if you want to run backups of certain datalists with
different protection policies, for example, if you may want to apply different
protection policies to different levels of backup (full, incremental, etc.). The
Override option frees you from having to create separate datalists for
different data protection schemes. To change the protection, select the
datalist you want to modify.
1. Open the main OmniBack II Backup window, and select the desired
datalist.
2. Choose Edit → Change Schedule. The Scheduler window appears.
4-34
Backing Up Your System
Setting Up Scheduled Backups
3. Choose Options → backup type. The Scheduler Option window
appears. In our example, we chose Full.
4. Click and drag on the Protection button to choose a protection period you
want to apply to the type of backup you selected in step 3. In our example,
full backups for this datalist will be protected for the period we select. The
date entry field appears.
5. In the date entry fields, enter the days, weeks or date until which you want
the data protected and click on the OK button to register the settings. The
window disappears and returns you to the Scheduler window.
To implement a schedule protection override using the command line, enter
the omnib command with the -datalist and -protect options and the
number of days, weeks, etc.
For example, enter: omnib -datalist datalistname -protect
weeks 2 . For more information, see manpages for the omnib command..
To Undo an Override
Select the datalist and the type of backup. In the dialog box of the
Scheduler Options window click on Default. This will cause OmniBack
II to use the protection originally specified in the datalist.
4-35
Backing Up Your System
Setting Up Scheduled Backups
4-36
5
Restoring Data
5-1
Restoring Data
In This Chapter
This chapter describes how to use the graphical user interface (GUI) to:
•
restore single files and directories from a filesystem backup
•
restore and search for files using the new simple restore GUI
•
restore single files from a rawdisk backup
•
restore disk sections from a rawdisk backup
The HP OmniBack II GUI is designed to make it easy to locate and restore
files and disks backed up with OmniBack II. It is designed to guide you
through the restore process, allowing you to select from lists whenever
possible. OmniBack II also offers a command-line interface for restores. For
more information on the command-line interface for restores, see the omnir
manpage.
5-2
Restoring Data
Restore Overview
Restore Overview
Basically, restoring data consists of providing the following information:
FROM
From which backup session to restore
WHAT
What files, directories, or disks to restore
HOW
Which restore options are to be used
TO
Where the files, directories, and disks are to be restored
OmniBack II helps you provide this information by guiding you through the
restore process.
OmniBack II provides three types of restores, depending on how the data
was backed up:
•
File restore from a filesystem backup
•
File restore from a rawdisk backup
•
Rawdisk restore from a rawdisk backup
A filesystem backup includes the logical structure of the data. Information
about this logical structure as well as the data are written to the media
management database. A rawdisk backup does not include information
about the structure of the data, therefore, the media management database
has no information about the individual files in the backup. Thus, the backup
type influences your ability to locate and restore individual files. For this
reason, each of these restore types is explained separately in the following
sections.
5-3
Restoring Data
Restoring Files from a Filesystem Backup
Restoring Files from a Filesystem Backup
The Media Management database of HP OmniBack II tracks the following
information about each backup performed using OmniBack II:
•
date and time of each backup
•
type of backup
•
directories, files, and disks that were backed up
•
file positions on the media
•
messages resulting from a backup or restore session
•
expiration date for the protection of the data contained on the backup
media
OmniBack II uses this information to help you locate and restore data.
Locating and restoring files and directories from a filesystem backup
consists of the following steps:
1. Selecting the backup object containing your file.
2. Specifying which files or directories to restore. You can use a File
Browser.
3. Specifying the Restore Options (optional)
4. Activating and monitoring the restore process.
All file restore tasks start from the OmniBack II Restore window. To
open this window, select Restore from the OmniBack II Main window.
Changing the view of backup objects
OmniBack II offers two types of views of backup objects:
Browser
5-4
offers a graphical representation of backup objects. This
view is suitable for backups with smaller number of
Restoring Data
Restoring Files from a Filesystem Backup
backup objects.
List
offers a text list of backup objects. This view is suitable
for backups with larger number of backup objects.
To change the view:
1. In the OmniBack II Restore window choose View → View
Objects As and select appropriate view.
Selecting the backup object
Figure 5-1
OmniBack II Restore Window
1. In the OmniBack II Restore window double click the backup object
which contains the files and directories you want to restore.The
FileSystem Restore Dialog window opens.
5-5
Restoring Data
Restoring Files from a Filesystem Backup
Selecting the files to restore
Figure 5-2
Browsing Files
Browsing Files to Restore
NOTE:
If you have chosen the No log backup options (no details on backed up files and
directories are stored in the database) or you have purged detail catalogs from the
database, you will not be able to use the browse functionality. Instead, OmniBack II
opens the Filesystem Restore Summary window where you type in the files
you want to restore.
See “Description of backup options” on page 4-22 and “Managing the Growth of the
Database” on page 8-4 for more information.
5-6
Restoring Data
Restoring Files from a Filesystem Backup
OmniBack II allows you to browse the list of files and directories you can
restore:
•
Double click on a directory in the FileSystem Restore Dialog window to
expand that directory, that is, opens other directories and files on that path.
•
Double clicking a second time on a directory above contracts the
directory, that is, closes directories and files on that path.
Shaded directories can only be expanded or contracted but not selected
for restore; this is because only a subtree of the directory was backed up,
for example: the /users directory is shaded because you only did a
backup of /users/mike/myfiles.
Manually typing files and directories
To manually enter the names of the files and directories you want to restore,
follow these steps:
1. In the FileSystem Restore Dialog window choose Restore
summary/details window... to open the FileSystem Restore
Summary window.
2. Enter the file or directory name in the field next to the Add button and
choose the Add button.
Repeat these steps to add additional files and directories.
Selecting a File/Directory Version
Generally, you have more than one backup containing the files you want to
restore, that is; you have several versions of a file or directory to choose to
restore from.
5-7
Restoring Data
Restoring Files from a Filesystem Backup
To Restore a Lost File
If a file has been lost, you will typically want to restore the most recent copy
of that file. The Files&Directories field in the FileSystem Restore Dialog
window by default displays the most recent backup of a file or directory.
1. Clicking the RIGHT mouse button on the required file (directory)
displays a cascading menu with the name of the selected file (directory)
as the title and the option to select either the Last Version, or a
Selected Version of the file for restore.
2. Clicking the RIGHT mouse button on Last Version automatically selects
the latest version of the file (directory) and highlights it in the FileSystem
Restore Dialog window.
3. Repeat the process to select further files or directories.
4. To deselect a file, click with the RIGHT mouse button on the file and
choose Deselect.
NOTE:
A shortcut to select/deselect the last version of a file (but not a directory) is to use the
Left mouse button.
To restore an older version of a file
If a file has been corrupted, you may want to restore an older version of the
file. In this case, recover the file from a backup containing an uncorrupted
version of the file. To select an older backup containing this file, proceed as
follows:
1. Select (highlight) the required directory or file displayed on the
FileSystem Restore Dialog window.
2. Click the RIGHT mouse button on the file (directory) and choose Select
Version to open the File Versions window which displays the available
versions of that file.
3. Select the desired version.
5-8
Restoring Data
Restoring Files from a Filesystem Backup
Specifying Fileset Restore Options
The fileset restore options apply to the specific files or directories selected
for restore in the FileSystem Restore Summary window.
1. After you have selected a version of the file (directory) to restore, choose
Restore summary/details window... to open the FileSystem Restore
Summary window.
2. Choose Options and enter the restore options you want to apply to the
selected files or directories. See below for a description of available
options.
Figure 5-3
Restore Options
Repeat the procedure to add options to filesets as needed.
Restore As
Restores the file or directory as the pathname
that you enter in the field to the right of the
5-9
Restoring Data
Restoring Files from a Filesystem Backup
button. The destination pathname can be an
existing pathname or a new pathname. This
enables you to rename objects as you restore
them
Restore Into
Restores the Fileset into the pathname that you
enter in the field to the right of the button. This
appends the fileset to the pathname. The
destination pathname must be an existing
pathname.
Match
Restores only files matching the enter file name
criterion. You can use wildcards to specify the
file name.
Exclude
Excludes the files and directories which you
enter in the field from the restore. Enter the full
pathname of the file. Wildcards are not allowed.
For example, you may want to exclude the
/tmp directory.
Skip
Excludes files matching the enter file name
criterion. You can use wildcards to specify the
file name. For example, you may want to skip
all *.o files in the selected fileset.
5-10
Restoring Data
Restoring Files from a Filesystem Backup
Selecting options for the restore session
These options apply to the whole restore session and all files (directories)
restored with one session.
1. Choose Options in the FileSystem Restore Dialog window to open
the Filesystem Restore Options window.
2. Enter the restore options you want to apply to the selected files or
directories. See below for a description of available options
Target Hostname
By default, files are restored to their original
hosts. You may want to choose some other host
in this field.
Merge
Merges existing files with files from the
backup. If a file on the disk is newer than the
backed up version, the newer file is kept. If a
file on the disk is older than the backed up
version, the file is overwritten with the new
version from the backup. The default is ON.
Overwrite
Replaces the files and directories on the disk
with the version of the files and directories on
the media. The default is OFF.
No Overwrite
Preserves the files and directories on the disk, if
they exist. The version of these files and
directories on the media is not restored. The
default is OFF.
Restore Protection
Attributes
Restore Time Attributes
Preserves the original protection attributes of
each restored file. When set to OFF, OmniBack
II applies the protection attributes of the current
restore session. The default is ON.
Preserves the time attribute values of each
restored file. When set to OFF, OmniBack II
sets the time attributes of the restored objects to
5-11
Restoring Data
Restoring Files from a Filesystem Backup
the current date and time. The default is ON.
Restore Sparse Files
Lock Files During
Restore
Restores sparse files in their original form. This
is important because sparse files can consume
additional disk space unless they are restored in
their original form. The default is OFF.
Denies access to files during the restore. By
default the option is set to OFF and the files are
not locked.
Omit Deleted Files
Deletes files which were removed between a
full and incremental backup. This option is
useful when you restore a complete directory or
disk and want to have only those files which
were in the directory or disk when the
incremental backup was done. The default is
OFF.
Move Busy Files
This option is used with the Omit Deleted Files
or Overwrite options. If the file is busy during
restore, OmniBack II cannot restore file. This
option moves the busy file filename to #filen
ame (with a hash in front of the filename.)
List Restored Objects
When set to ON, OmniBack II displays the
names of the files and directories in the
Message section of the Restore Monitor
window as the objects are being restored. The
default is OFF.
Display Statistical
Information
Pre- exec command
5-12
When set to ON, OmniBack II displays detailed
statistical information about the restore session.
The default is OFF.
A Pre- exec command is executed as soon as
the Disk Agent is started, but before any restore
processing has started. See “Using Pre- and
Post- exec functions” on page 7-6 for more
information.
Restoring Data
Restoring Files from a Filesystem Backup
Post- exec command
A Post- exec command is executed just before
the Disk Agent terminates. See “Using Pre- and
Post- exec functions” on page 7-6 for more
information.
Activating and monitoring the restore
Before you actually run the restore session, you may want to preview it.
The Restore Monitor window also allows you to change to a different
logical device for restore or to preview a restore session.
Figure 5-4
Monitoring the Restore Session
To activate the restore, follow these steps:
1. After you have entered and selected all the requested information needed
for the restore in the Filesystem Restore window, click the OK
button. The Restore Monitor window appears with the device status
indicating the Inactive/Waiting status and the session indicating
5-13
Restoring Data
Restoring Files from a Filesystem Backup
Pending.
2. Choose:
•
Actions → Preview Restore to preview the restore session.
•
Actions → Start Restore to start the restore session.
3. Select the Network Load.
Setting the option to Low reduces the load on the network when running
OmniBack II. This prevents the data transmission for restore from
blocking the network for other users, but increases the time required for
the restore.
4. Select Report Level.
By default all messages are displayed - the warning level. If you select
any other levels, only messages with higher priority are displayed.
5. Click on Start Restore to start running the restore.
The Status field indicates the session is Pending then Running. If
OmniBack II requires media, Mount Request appears in the Status
field.
To Respond to Mount Requests
When prompted by a Mount Prompt message from the Restore Monitor,
insert the correct media in the device and confirm the prompt request.
1. Close the Mount Request dialog.
2. Select (highlight) the device with the Mount Request status.
3. Select Actions → Confirm Mount Request from the menu bar.
The status of the session and device changes to Running. When the
session has completed, the status of the session displays Completed and
a window appears with the message "Session Completed!".
6. Click OK to close the message window and return to the OmniBack II
Restore Monitor window.
7. Select File → Close to close the monitor window and return to the
OmniBack II Restore window.
5-14
Restoring Data
Restoring Files from a Filesystem Backup
Viewing the Messages Field
While the restore is running, messages appear in the Restore Monitor
window Messages field indicating the status of the filesets, devices and the
progress of the restore.
While messages are still displayed in the Messages field, position the
cursor in the Messages field and click on the RIGHT mouse button. Choose
from the following message options in the menu that appears:
Search
To search for specific string in the message line.
Clear
To clear all messages from the field.
Scroll Messages To scroll messages as they are generated.
Print Messages To print messages on the printer.
The maximum buffer for messages in the Message field is limited with the
configurable parameter MAXGUIMSG. See “OmniBack II Global Options
File” on page 7-11 for more information. If the maximum buffer is
exceeded, messages stop appearing in the Message field and the error
message “Output disabled” appears, but the backup continues.
If the maximum buffer is exceeded, messages stop appearing in the
Message field, and the error message, Output disabled appears
although the restore continues.
Searching for a specific file
From time to time you may forget where a specific file was located on your
system, although you remember a part of the file name. OmniBack II gives
you a way to easily search for such files.
Note that you should normally use the browser to find your files. The search
action may take longer and it will only show the first 500 files matching the
selected criteria. In the event this limit is reached, you must further limit
your search criteria.
1. From the OmniBack II Main menu choose Restore to open the HP
OmniBack II Restore window.
5-15
Restoring Data
Restoring Files from a Filesystem Backup
2. Choose Actions → Search… to open the HP OmniBack II Restore
- Search window.
3. Enter as much information as possible to speed up the search. Press
Search to start searching for the file. Note that this action can take some
time, depending on your configuration.
4. If you want to restore all files with the same name, select an object from
the File Found In list and click OK to open the FileSystem
Restore Dialog window. Continue with the normal restore procedure.
5. If you want to select a specific file, double click on the host in the File
Found In list and select the specific file. Continue with the normal
restore procedure.
5-16
Restoring Data
Restoring a Rawdisk Section (Logical Volume)
Restoring a Rawdisk Section (Logical Volume)
Restoring a disk section (logical volume) includes the following tasks:
1. Selecting from which backup session you want to restore.
2. Selecting a version of backup.
3. Selecting a section to restore.
4. Selecting restore options.
5. Previewing, starting and monitoring the restore.
5-17
Restoring Data
Restoring a Rawdisk Section (Logical Volume)
Choosing the section to restore
1. Click Restore at the OmniBack II Main Menu to open the OmniBack
II Restore window.
2. Select the rawdisk backup from which you want to restore.
3. Select Actions → Restore Selected… to open the Rawdisk Restore
Dialog window.
4. In the Backed Up On field choose the version you want to restore. Click
... to get a list of versions.
5. The Sections field shows all disk sections backed up in the selected
backup session. To remove a disk section from the list, select the section
and click the Remove button. If you need to add it back, type its name in
the field below and click Add.
Figure 5-5
Restoring a Rawdisk
5-18
Restoring Data
Restoring a Rawdisk Section (Logical Volume)
Specifying rawdisk restore options
OmniBack II provides restore options which allow you to control the restore
session.
To select the restore options, choose Rawdisk Restore Options.
The following options are available:
Target Hostname
Allows you to change the hostname where the
rawdisk section should be restored. By default,
OmniBack II restores to the original system.
Pre- exec
A Pre-exec command is executed as soon as the
Disk Agent is started, but before any restore
processing has started.
Post- exec
A Post-exec command is executed just before
the Disk Agent terminates.
For more information see “Using Pre- and Postexec functions” on page 7-6.
List Restored Objects
Display Statistical
Information
When set to ON, OmniBack II displays the
names of the disk sections in the Message
section of the Monitor window as the objects
are restored. The default is OFF.
When set to ON, OmniBack II displays detailed
information about the restore session. The
default is OFF.
Starting and monitoring a rawdisk restore
When you activate the restore, you open the OmniBack II Restore
Monitor window, enabling you to monitor the progress of your restore. The
Monitor window also allows you to respond to mount requests for the
restore. To activate the restore, follow these steps:
5-19
Restoring Data
Restoring a Rawdisk Section (Logical Volume)
1. From the Rawdisk Restore Dialog choose OK to open the Restore
Monitor window.
2. Choose:
•
Actions → Preview Restore to preview the restore session.
•
Actions → Start Restore to start the restore session.
3. Select the Network Load.
Setting the option to Low reduces the load on the network when running
OmniBack II. This prevents the data transmission for restore from
blocking the network for other users, but increases the time required for
the restore.
4. Select the Report Level.
By default all messages are displayed - the warning level. If you select
any other levels, only messages with higher priority are displayed.
5. Click on Start Restore from the menu bar to start the restore. The Status
field indicates the restore session is Inactive/Waiting, then finally
Running. Other status messages may appear depending on the progress
of the restore.
When the session has completed, the status of the session displays
Completed and a window with the message "Session Completed!"
appears.
Responding to mount requests
A mount prompt is issued when OmniBack II needs new media to read data.
1. Insert needed media in the device.
2. Select the device with the Mount Request status.
3. Select Actions → Confirm Mount Request from the menu bar. The
status of the session and device changes to Running.
5-20
Restoring Data
Restoring Files from a Rawdisk Backup
Restoring Files from a Rawdisk Backup
Restore of files from the rawdisk backup is supported on HFS filesystems
only.
You can restore only "Regular" Unix files out of the rawdisk backup.
Recovery of any other file types is not supported by OmniBack II.
You can restore files from any directory except the CDF directory (of cluster
environments.)
5-21
Restoring Data
Restoring Files from a Rawdisk Backup
Selecting the file to restore
1. Click Restore at the OmniBack II Main Menu. The OmniBack II
Restore window appears.
2. Select the rawdisk backup object which contains the file you want to
restore.
3. Select Actions → Restore Single File... to open the Single File
Restore window. The backup object you selected is displayed in the
Rawdisk field.
Figure 5-6
Restoring Single Files from a Rawdisk
4. In the Backed Up On field choose the version from which you want to
restore. Click ... to get a list of versions.
5. In the Restore File enter the path and a file name.
5-22
Restoring Data
Restoring Files from a Rawdisk Backup
6. Enter the disk section name in the field labeled From Section, or click
the ... button to select the section from a list.
Specifying rawdisk restore options
The rawdisk restore options apply to all files to be restored.
1. Choose Single File Restore Options ... to open the Filesystem
Restore Options window.
The following options are available:
Target Hostname
Allows you to change the hostname where the
rawdisk section should be restored. By default,
OmniBack II restores to the original system.
Merge
Preserves the destination version of files and
merges the restore fileset into the existing
fileset on the disk. The default is Merge.
Overwrite
Replaces the files and directories on the disk
with the version of the files and directories on
the media.
No Overwrite
Preserves the files and directories on the disk, if
they exist. The version of these files and
directories on the media is not restored.
Restore Time Attributes
Preserves the time attribute values of the
restored file. When set to OFF, OmniBack II
sets the time attributes of the restored file to the
current date and time. The default is ON.
Restore Protection
Attributes
List Restored Files
Preserves the original protection attributes of
the restored file. When set to OFF, OmniBack II
applies the protection attributes of the current
restore session. The default is ON.
When set to ON, OmniBack II displays the
name of the file in the Message section of the
5-23
Restoring Data
Restoring Files from a Rawdisk Backup
Monitor window while the objects is being
restored. The default is OFF.
Display Statistical
Information
When set to ON, OmniBack II displays detailed
information about the restore session. The
default is OFF.
Pre- exec
A Pre-exec command is executed as soon as the
Disk Agent is started, but before any restore
processing has started. See “Using Pre- and
Post- exec functions” on page 7-6 for more
information.
Post- exec
A Post-exec command is executed just before
the Disk Agent terminates. See “Using Pre- and
Post- exec functions” on page 7-6 for more
information.
Starting and monitoring file restore from a rawdisk
After you have provided the information necessary for the restore, you are
ready to activate the restore.
1. In the Single File Restore window choose OK to open the
OmniBack II Session Monitor window.
2. Choose:
•
Actions → Preview Restore to preview the restore session.
•
Actions → Start Restore to start the restore session.
3. Select the Network Load.
Setting the option to Low reduces the load on the network when running
OmniBack II. This prevents the data transmission for restore from
blocking the network for other users, but increases the time required for
the restore.
5-24
Restoring Data
Restoring Files from a Rawdisk Backup
4. Select the Report Level.
By default all messages are displayed - the warning level. If you select
any other levels, only messages with higher priority are displayed.
5. Click on Start Restore from the menu bar to start the restore. The Status
field indicates the restore session is Inactive/Waiting, then finally
Running. Other status messages may appear depending on the progress
of the restore.
When the session has completed, the status of the session displays
Completed and a window with the message "Session Completed!"
appears.
Restoring a file when you do not know the filename
If you need to restore a file or directory but do not know the filename, do the
following:
1. restore the entire rawdisk to an empty section
2. Mount this empty section as a filesystem.
3. Find the file you are looking for.
4. Copy the file back to its original location.
5. Umount the empty section.
5-25
Restoring Data
Restoring Files from a Rawdisk Backup
5-26
6
Monitoring Sessions
6-1
Monitoring Sessions
In This Chapter
OmniBack II allows you to monitor currently running sessions, as well as
viewing already finished session. You can view messages and status of the
sessions, devices used, objects backed up or restore, and so forth.
Beside that, a Cell Monitor is provided that helps administrators of
environments with more cells to have an overview of all their cells.
6-2
Monitoring Sessions
Monitoring Running Sessions
Monitoring Running Sessions
When you run an interactive backup or restore session, a monitor window is
opened showing objects, devices, and messages generated in the session.
Even if the GUI is closed, the sessions continue.
OmniBack II allows you to connect to any running backup session and
monitor the progress of the session, abort the session, or respond to a mount
prompt request. For more information on mount prompt requests see
“Automating Mount Prompt Handling” on page 7-3.
A progress and status of the running sessions can be monitored using the
OmniBack II Session Monitor window. As soon as a session is
finished or is aborted it is no longer displayed as the currently running
session but as the previous session.
1. From the OmniBack II Main menu choose Monitor to open the
OmniBack II Session Monitor window.
The window shows the progress and the status of all currently running
session in a cell.
From the Actions menu you can abort a session.
Viewing details of a currently running session
You can view detailed information of a currently running session by double
clicking on the session to open the OmniBack II Backup Monitor or
OmniBack II Restore Monitor window.
6-3
Monitoring Sessions
Monitoring Running Sessions
Figure 6-1
Monitoring a Session
In this window you can see all backup or restore objects with their status,
status of devices, and messages generated in this session.
From the Actions menu you can abort the session or cancel the mount
prompt request.
By clicking the right mouse button in the Messages area you can scroll,
clear, print, or search for specific messages.
6-4
Monitoring Sessions
Viewing Previous Sessions
Viewing Previous Sessions
As soon as a session is finished or is aborted it is no longer displayed as a
currently running session, but as a previous session.
1. From the OmniBack II Main menu choose Monitor to open the
OmniBack II Session Monitor window.
2. Choose View → Previous Sessions and choose all or selected sessions.
The window shows all completed and currently running sessions.
Figure 6-2
Viewing Completed Sessions
The following actions are available from the OmniBack II Session
Monitor window showing previous sessions:
1. Restarting unsuccessful backups. See “Managing Failed Backups” on
page 7-12 for more information.
2. Removing detailed information from the session. See “Managing the
Growth of the Database” on page 8-4 for more information.
6-5
Monitoring Sessions
Viewing Previous Sessions
Viewing details on previous sessions
You can view detailed information of a previous session by double clicking
on the session to open the OmniBack II Backup Monitor or OmniBack
II Restore Monitor window.
In this window you can see all backed up or restored objects with their
status, as well as messages generated in this session. Using the View menu
you can also find media used in this session. This functionality is available
for backup sessions only.
By clicking the right mouse button in the Messages area you can scroll,
clear, print, or search for specific messages.
By double clicking on a specific object you can see only messages related to
this object.
6-6
Monitoring Sessions
Monitoring Several Cells
Monitoring Several Cells
For environments with more OmniBack II cells configured, a special
monitor is provided giving you an overview of all your cells.
To start the Cell Monitor, type:
xomnicellmon -cell Cell_Server1 Cell_Server2 ...
You must specify a list of Cell Server systems of the cells you want to
monitor.
Figure 6-3
Monitoring OmniBack II Cells
From the Actions menu you can view different types of backup objects,
lists scheduled objects, or list disks not configured for backup.
6-7
Monitoring Sessions
Monitoring Several Cells
6-8
7
OmniBack II Advanced Tasks and Concepts
7-1
OmniBack II Advanced Tasks and Concepts
In This Chapter
This chapter describes the customization features and provides instructions
for configuring these features.
The following customization features are described in this chapter:
•
Automating Mount Prompt Handling
•
Pre- and Post-Exec Processing of Commands
•
Customizing the Encryption Algorithm
•
Customizing the Compression Algorithm
•
Customizing the OmniBack II Global Options
•
Recovery after a Crash
•
Moving a Cell Server
•
Recover Failed Backups
7-2
OmniBack II Advanced Tasks and Concepts
Automating Mount Prompt Handling
Automating Mount Prompt Handling
The Mount Request status on the monitor screen indicates that the device
needs a new medium and is waiting for a user response to the mount request.
OmniBack II remains in this state until the mount request is confirmed or the
session is aborted. If there is no responds to the mount prompt within a
predefined mount prompt delay (30 minutes by default), the mount prompt
script is executed on the Cell Server which sends an email to the user who
started the session.
The default script that sends email to the user is:
/usr/omni/bin/Mount.sh (on HP-UX 9.0x)
/opt/omni/lbin/Mount.sh (on HP-UX 10.x)
The Mount Prompt Delay and the Mount Prompt Script can be configured
using Advanced Options in the Logical Device Configuration
window.
You can change the default script by modifying the MountScript variable
in the global options file.To change the default mount prompt delay, modify
the MountDelay variable in the global options file.
OmniBack II invokes the mount prompt script and provides the following
command line arguments:
1.
2.
3.
4.
5.
6.
7.
8.
Unix login of user who started the backup
Unix group of user who started the backup
host where the backup was started
PID of the backup process
session key
logical device that generated the mount request
host of the logical device
pathname of the physical device name waiting for new medium (for
exchangers, the slot number (SCSI Exchanger) or the device file (MO
7-3
OmniBack II Advanced Tasks and Concepts
Automating Mount Prompt Handling
Jukebox) is used instead of the pathname of the physical device)
9. device type number
10. device type name
11. medium id
12. medium label
13. medium location
14. pool name
15. media allocation policy of the pool
16. media type number
17. media type name
18. session key
If OmniBack II is unable to define a medium to mount, empty parameters
are sent to the mount prompt script. This can occur if there are no media
remaining in the pool or if there is no space remaining on the medium.
For more information on adding a script file, see the online help topic
"Configuring Advanced Options for a Logical Device" in "Logical Device
Configuration Tasks".
Mount prompt script examples
This section provides some typical modifications to the mount promt script.
An email should be sent to a specific user only:
In the Mount.sh script change the $USER@$HOSTNAME variable to the
specific user.
Aborting sessions with the mount prompt script:
Specify the following commands in the mount prompt script.
On HP-UX 9.0x:
/usr/omni/bin/omniabort -session $SESSIONKEY 2>/dev/null
1>/dev/null
7-4
OmniBack II Advanced Tasks and Concepts
Automating Mount Prompt Handling
On HP-UX 10.x:
/opt/omni/bin/omniabort -session $SESSIONKEY 2>/dev/null
1>/dev/null
Enter all commands in one line only.
7-5
OmniBack II Advanced Tasks and Concepts
Using Pre- and Post- exec functions
Using Pre- and Post- exec functions
Before a backup can begin, additional actions are sometimes necessary. For
example, before a backup begins, transaction processing on a database must
stop and the users must be notified.
Pre- and Post- exec actions are provided on two levels:
•
Before and after the backup of each datalist (session) is started. These
actions are executed on the Cell Server only.
•
Before and after the backup of each object in a datalist is started. These
actions are started on client systems where the object resides.
Pre- and Post- exec functions are implemented using the Unix pipe
mechanism. All processes started in the Pre- or Post- exec functions must
finish before processing continues.
NOTE:
Make sure that you follow instructions described in the following sections when
creating your Pre- and Post- exec scripts.
Pre- and Post- exec command for a datalist
Pre- and Post- exec command for a datalist are started before and after the
backup of a datalist (session). These commands are executed on the Cell
Server only.
The following environment variables are exported and can be used in Preand Post- exec scripts for a datalist (session) on the Cell Server only:
SESSIONKEY Is used to identify the running session. You may, for
example, abort a backup session before it is started if
something is wrong. You can use this variable with Pre
exec commands only.
PREVIEW
7-6
Set to 1 if preview is running. Set to 0 if backup is
running. You can therefore modify your commands to be
OmniBack II Advanced Tasks and Concepts
Using Pre- and Post- exec functions
executed only during a backup and not preview.
SESSIONID
Is used to identity a finished session and is recorded in the
database. You can only use this variable with the Postexec command.
SMEXIT
Exit code of the Session Manager and is the same as the
exit code of the omnib command:
•
•
•
•
•
0 Completed
10 Completed with file errors
11 One or more Disk Agents failed
12 All Disk Agents failed
13 Session was aborted
You can only use this variable with the Post-exec
commands.
Make sure that you check the following before configuring Pre- and Postexec commands for a datalist:
•
If a Pre- exec command fails (returns a non-zero value), the backup status
of the session is set to failed and the session is aborted.
•
If a Post- exec command fails (returns a non-zero value), the backup of
the session is set to completed with errors.
•
The Pre- and Post- exec commands for a datalist are by default NOT
executed during a preview of the backup. This behavior is defined by the
ExecScriptOnPreview variable in the global options file. See
“OmniBack II Global Options File” on page 7-11 for more information.
•
Pre- and Post- exec commands are handled in the same way as commands
entered at the shell prompt. Special shell characters such as the pipe | and
the redirect symbols > and < are allowed.
•
The Unix shell controls the Pre- and Post- exec processing. Therefore, you
cannot use OmniBack II (GUI or command-line) to stop a session. The
only way to stop a session while it is executing a Pre- or Post- exec
command is to use the Unix kill command.
•
The Pre- and Post- exec processes operate in background mode.
Therefore, do not use any interactive commands for Pre- and Post- exec
7-7
OmniBack II Advanced Tasks and Concepts
Using Pre- and Post- exec functions
processing.
•
If a command writes any text to stdout, this text is received by the Disk
Agent, sent to the Session Manager, and written to the database. The
stderr is redirected to /dev/null. You can redirect it to stdout to get
error messages logged to the database.
•
The Pre- and Post- exec commands for a datalist must be located on the
Cell Server.
•
All commands are started with the UID of the user configured as the
session owner. If an user is not configured on the Cell Server, Pre- and
Post- exec commands are not executed but the backup continues.
•
The Pre- and Post- exec scripts must send some output at least every 15
minutes by default or the scripts get aborted. You can change this time
interval by modifying the ScriptOutputTimeout variable in the global
options file.
•
By default, Pre- and Post- exec functions can be located in any directory
on the Cell Server. If you set the UsePanScripts variable in the global
options file, scripts can be started from a specific directory only. See the
global options file for details.
Pre- and Post- exec commands for an object
Pre- and Post- exec commands for an object are executed before the backup
of the object is started or finished. You can specify these commands for each
object independently. These commands are executed on the system where
the Disk Agent is running.
The following environment variables are exported and can be used in the
Pre- and Post- exec scripts for an object on the system where the Disk Agent
is running:
PREVIEW
Set to 1 if preview is running. Set to 0 if backup is
running. You can therefore modify your scripts to be
executed only during a backup and not during preview.
BDACC
The Disk Agent sets its exit code (zero is successful) to
the BDACC environment variable. This variable can be
7-8
OmniBack II Advanced Tasks and Concepts
Using Pre- and Post- exec functions
checked in the Post- exec script thus making the Postexec command dependent on the successful termination
of the Disk Agent.
Make sure that you check the following before configuring Pre- and Postexec commands:
•
The Pre- and Post- exec commands for an object are also executed during
the preview of a backup. Therefore, you may want to preview your
backup first and then add the Pre- and Post- exec commands, or check the
PREVIEW environment variable in your scripts.
•
If a Pre- exec command for an object fails (returns a non-zero value), the
backup status of the object is set to aborted and the Disk Agent stops
processing. Therefore, no backup of the object exists.
•
If a Post- exec command fails (returns a non-zero value), the backup status
of the object is set to aborted. However, the backup of the object exists and
data can be restored.
•
The Pre- and Post- exec commands should sent some output to the Disk
Agent at least every 120 minutes by default, or the backup of this object
will be aborted. This timeout period can be changed by modifying the
SmDaIdleTimeout variable in the global options file.
•
Pre- and Post- exec commands are handled in the same way as commands
entered at the shell prompt. So, special shell characters such as the pipe |
and the redirect symbols > and < are allowed.
•
The Unix shell controls the Pre- and Post- exec processing. Therefore, you
cannot use OmniBack II (GUI or command-line) to stop the Disk Agent.
The only way to stop a session while it is executing a Pre- or Post- exec
command is to use the Unix kill command.
•
The Pre- and Post- exec processes operate in background mode.
Therefore, do not use any interactive commands for Pre- and Post- exec
processing.
•
If a command writes any text to stdout, this text is received by the Disk
Agent, sent to the Session Manager, and written to the database. The
stderr is redirected to /dev/null. You can redirect it to stdout to get
error messages logged to the database.
•
The Pre- and Post- exec commands for an object must be located on the
7-9
OmniBack II Advanced Tasks and Concepts
Using Pre- and Post- exec functions
system where the Disk Agent is running.
•
The Pre- and Post- exec commands must be executable and should be
specified with the full pathname.
•
You can override any Pre- or Post- exec commands specified for objects
or datalists by specifying them with the OB2PREEXEC and
OB2POSTEXEC variables in the global options file.
7-10
OmniBack II Advanced Tasks and Concepts
OmniBack II Global Options File
OmniBack II Global Options File
OmniBack II uses the Global Options file for advanced configuration
settings. Normally you do not need to modify this file, but occasionally you
can achieve different functionality if you modify some parameters. The
description of all variables is in the file itself.
The global options file /usr/omni/config/options/global (on HPUX 9.0x) or /etc/opt/omni/options/global (on HP-UX 10.x): is an
ASCII file. Using a text editor you need to uncomment the appropriate
option and set the value.
NOTE:
When you update to the new version of OmniBack II, the global options you have are
not overridden. The new global options file is saved in the newconfig directory
tree and new options can be merged manually.
7-11
OmniBack II Advanced Tasks and Concepts
Managing Failed Backups
Managing Failed Backups
Recovering a failed backup is closely related to how OmniBack II runs full
and incremental backups.
Incremental backups always depend on the last full backup. In there is no
full backup, OmniBack II will perform one even if an incremental backup is
scheduled - you cannot have an incremental backup without a prior full
backup.
Failed system backup
During a backup operation, it can happen that a system is down or
unavailable due to maintenance or problem reasons. This system will not be
backed up.
•
If a full protected backup exists, an incremental backup will be started
next time the system is scheduled for backup.
•
If a full protected backup does not exist, a full backup will be started next
time the system is scheduled for backup.
OmniBack II therefore ensures that the backup is always as current as
possible.
Note that if you have configured infrequent scheduled backups, there may be
a period of time when a recent backup for a system does not exist.
You can avoid this situation by running a backup for the system interactively
when systems are available. See “Running Interactive Backups” on
page 4-14 for instructions.
7-12
OmniBack II Advanced Tasks and Concepts
Managing Failed Backups
Restarting failed backups
During a backup session it may happen that some backup objects are not
backed up successfully. OmniBack II provides a simple way to restart the
backup of failed objects.
Sessions with status running or completed or backup of a barlist cannot be
restarted.
Using GUI:
1. From the Monitor window choose View → Previous Sessions.
2. Select a failed session and choose Actions → Restart Unsuccessful
Backup.
This starts the backup of all failed objects.
Using commands:
For example, you want to restart failed objects from the datalist MIKI.
1. To get a list of all objects in the backup session, you must first find the
session ID of the datalist backup session.
$ omnidb -session -datalist MIKI -latest
SessionID
Type
Status
User.Group@Host
==================================================================
1996/01/08-12
Backup
Completed/Failure root.sys@ik01.rep.com
2. To list all objects and their status of the backup session:
$ omnidb -session 1996/01/08-12
Object Name
Object Type
Object
Status
==================================================================
kpp3.rep.com:/users1 'REP'
FileSystem
Completed
kpp2.rep.com:/users/frenk 'REP'
FileSystem
Completed
mdd56.rep.com:/users 'REP'
FileSystem
Failed
trblexll.rep.com:/home 'REP'
FileSystem
Failed
waikik.rep.com:/C 'REP'
FileSystem
Completed
3. To selectively restart failed objects only:
$ omnib -datalist MIKI -select mdd56:/users 'REP' trblexll:/home
'REP'
7-13
OmniBack II Advanced Tasks and Concepts
Detection of CDF files
Detection of CDF files
OmniBack II can detect CDF (Context Dependent Files) on cluster server
systems. Since this often consumes time, you can disable this detection to
improve overall backup performance, especially for large filesystems with a
number of small files.
Detection of CDF files can be selectively enabled/disabled on a per-host
basis.
On HP-UX 10.x, the CDF detection/backup is disabled by default, as the
operating system no longer supports CDFs.
On HP-UX 9.0x, the CDF detection/backup is by default enabled. It can be
disabled on a specific system by specifying the following line in the
/usr/omni/.omnirc file on that system:
OB2NOCDF=1
Note that ALL backup Disk Agents running on that system will receive this
option and disable the CDF detection/backup.
We recommend disabling the CDF detection/backup on all systems in acell
EXCEPT on servers of diskless cluster systems or any other systems that are
likely to have CDF files.
7-14
OmniBack II Advanced Tasks and Concepts
OmniBack II and Journaled File Systems on HP-UX 10.x
OmniBack II and Journaled File Systems on HPUX 10.x
OmniBack II supports JFS (VxFS) online backup feature.
For more information on JFS functionality see your system administration
manuals.
Concepts
JFS allows you to back up a filesystem while it is being used by some other
application. This is called an online backup and is done by creating a
snapshot of the filesystem and backing up this snapshot.
A snapshot of the filesystem is created when you mount the JFS filesystem
to a temporary directory. At this point you also specify the filesystem you
want to snap.
You configure a backup of this temporary directory, which actually is a
mount point to the snapshot of the filesystem as it was at the moment of
mount.
When the backup is finished, you unmount the snapshot filesystem so that it
can be used for other purposes.
Note that you can perform normal backups without using the JFS snapshot
feature by simply configuring backup as for any other filesystem.
If you want to use the JFS snapshot feature, you must configure a backup as
follows.
1. You must have an empty or unused partition created on your system that
can be used by JFS for a snapshot. See your system administrator’s
manual for instructions.
2. Create a temporary directory to which you will mount the snapshot
filesystem.
7-15
OmniBack II Advanced Tasks and Concepts
OmniBack II and Journaled File Systems on HP-UX 10.x
3. Create shell scripts to mount and unmount the snapshot filesystem to the
temporary directory. See “Script templates” on page 7-17 for templates of
these scripts.
4. Configure a backup of the temporary directory. The mount script must be
specified as the Pre- exec command, and the unmount script as the Postexec command.
Limitations
1. It is possible to restore JFS to HFS and HFS to JFS, except in the
following cases:
•
JFS does not support ACLs, therefore on restores from HFS to JFS this
information will be lost.
•
With JFS it is possible to define different extent attributes and
allocation controls for each file. OmniBack II recognizes and stores
this information and restores it back during restore. When restoring to
an HFS, this information is lost.
2. You cannot use the JFS snapshot feature on ordinary UNIX filesystems.
3. The amount of disk space necessary for the snapshot of a filesystem
depends on how busy the snapped filesystem is while the backup is run.
The recommended size is up to 15% of the snapped filesystem, if the
filesystem is in heavy use during the backup. Normally, the size will be
around 5%.
If the amount of data modified on the snapped filesystem is higher than
space available, OmniBack II produces Cannot stat error messages
for all the remaining files to be backed up. You should unmount the
snapshot filesystem and repeat the backup procedure.
7-16
OmniBack II Advanced Tasks and Concepts
OmniBack II and Journaled File Systems on HP-UX 10.x
Restoring
Since you have backed up the temporary directory with the snapshot
filesystem mounted and not the original directory, you must restore the files
to the original directory using the Restore As option.
Script templates
Here is a template that can be used to mount the JFS filesystem.
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
SnapMount.sh
Mounting snapshot filesystem (pre-exec script)
A script requires 3 parameters:
1. a block special file of the snapped FS
or
a mount point directory of the snapped FS
2. a block special file of the snapshot FS
3. a mount point of the snapshot FS
NOTE:
In case of multiple Disk Agents reading from the same snapshot FS,
the pre-exec script should contain a kind of synchronization
mechanism for following reasons:
1) an attempt to mount an already mounted snapshot FS,
snapping the same FS
will cause the pre-exec script to fail and a DA to abort
2) an attempt to mount an already mounted snapshot FS,
snapping some other FS will cause a warning to be generated,
script to fail and a DA to abort
3) a synchronization with the post-exec script should be also
provided because the snapshot FS must not be unmounted while there
is other DA reading from the FS.
SNAPPED_FS=$1
SNAPSHOT_FS=$2
MOUNT_POINT=$3
mount -F vxfs -e -o snapof=$SNAPPED_FS $SNAPSHOT_FS $MOUNT_POINT
#
# end SnapMount.sh
#
7-17
OmniBack II Advanced Tasks and Concepts
OmniBack II and Journaled File Systems on HP-UX 10.x
exit $?
Here is a template that can be used to unmount the JFS system.
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
SnapUnmount.sh
Unmounting snapshot filesystem (post-exec shell script)
Script requires 1 parameter:
- a mount point directory of the snapshot FS
or
- a block special file of the snapshot FS
NOTE
In case of multiple Disk Agents reading from the same snapshot FS,
a kind of synchronization mechanism has to be added for
the following reasons:
1) post-exec script should unmount snapshot FS only if there
is no other DA reading from the snapshot FS
Success/failure of the DA can be checked by examining
the BDACC environment variable
MOUNT_POINT=$1
umount -v $MOUNT_POINT
#
# end SnapUnmount.sh
#
exit $?
7-18
OmniBack II Advanced Tasks and Concepts
Deleting a session
Deleting a session
HP OmniBack II keeps track of how long data on the medium is protected
(see “Using Backup Options” on page 4-19 for more information.) When the
protection expires, the session will be overwritten by the newer backups.
OmniBack II provides a way to delete unwanted sessions from the database.
1. From the Monitor window choose View → Previous Sessions.
2. Select a session you want to delete and choose Actions → Purge Session.
The session disappears from the list.
The following happens:
•
Protection is removed from the session. This means that the session will
be overwritten.
•
All detail information about the session is removed from the database.
•
The session itself remains on the media until it is overwritten by the next
backup. Therefore, you can still see or restore the session using list from
media.
7-19
OmniBack II Advanced Tasks and Concepts
Recovery after a Crash
Recovery after a Crash
There are few simple rules you can follow to make sure that crash recovery
will be as smooth and simple as possible:
•
Make a backup of your Cell Server files and the files of other systems on
a regular basis. This ensures that you can always restore up-to-date data
and thus allow users to continue their work normally.
•
Make a full disk backup of all important systems in your cell on a regular
basis. This allows a quick recover from a disk crash.
•
Make a database backup each night to ensure that you can restore the last
state of the database. If possible, configure the database backup on a
separate logical device, for example, standalone DAT. This simplifies
recovery of files in the event of a Cell Server crash.
Preparing for a crash
A disk crash of any system is probably the worst thing that can happen,
therefore it is good to plan a recovery procedure for all systems. Basically,
there are three different crash cases:
•
Some files have been corrupted or deleted, but the system is still operating
normally. In this case you only need to restore the corrupted or missing
files. You can do this easily using OmniBack II restore.
•
A client system has suffered a major disk crash and is no longer operating
properly. See “Recovering a client system” on page 7-21 for instructions.
•
A Cell Server system has suffered a major disk crash and is no longer
operating properly. See “Recovering a Cell Server” on page 7-22 for
instructions.
7-20
OmniBack II Advanced Tasks and Concepts
Recovery after a Crash
Recovering a client system
When you need to recover a root disk of your system, you may follow one of
these two methods:
1. Connect a new disk on another system and restore the complete backup
on the disk. This is the preferred solution.
2. Install a new disk on the crashed system and restore files on the disk.
Both methods are explained below.
Recovery on another system
1. Connect a new disk on another system.
2. Make the disk bootable.
3. Restore your full disk backup on the new disk.
4. Move the disk to the crashed system and reboot the system.
Recovery on the same system
1. Replace the crashed disk with a new one.
2. Boot your system from the installation media for your system, such as a
DAT tape.
3. Reinstall the operating system. Refer to your system administrator’s
manual for instructions.
4. Reinstall OmniBack II client modules as they were before the disk crash.
See Chapter 2, “Installation Overview” on page 2-1 for instructions.
5. Reinstall other applications using installation media.
6. Restore other directories and files on the system from your backup. See
Chapter 5, “Restoring Data” on page 5-1 for instructions.
Your client system should be successfully recovered.
7-21
OmniBack II Advanced Tasks and Concepts
Recovery after a Crash
Recovering a Cell Server
Recovering of a Cell Server after a major disk crash includes some
additional steps. This procedure is also applicable in case your complete
Cell Server isdestroyed by some accident.
When you need to recover a root disk of your system, you may follow one of
these two methods:
1. Install the new disk to another system and restore required files on the
disk. This is a preferred solution.
2. Install the new disk on the crashed system and restore files on the disk.
Both methods are explained below.
Recover on another system
1. Install the new disk on another system.
2. Make the disk bootable.
3. Restore your full disk backup of the Cell Server on the new disk.
4. Restore the last database backup.
5. Move the disk to the crashed Cell Server system and reboot the system.
Recover on the same system
1. Replace the crashed disk. If you must install a new Cell Server, configure
it with the same name as the original Cell Server.
2. Boot your system from the installation media for your system, such as
DAT tape. You do not need to execute this step if you are installing a new
Cell Server.
3. Reinstall the operating system. Refer to your system administrator’s
manual for instructions. You do not need to execute this step if you are
installing a new Cell Server.
4. Reinstall the OmniBack II software on the Cell Server. See the HP
OpenView OmniBack II Installation Guide for instructions.
5. Restore the last backup of your database. Note that you cannot restore the
7-22
OmniBack II Advanced Tasks and Concepts
Recovery after a Crash
database directly. See “Restoring the Database” on page 8-14 for
instructions. This simplifies the restore of all other files from media.
6. Stop running all OmniBack II processes with the omnisv.sh -stop
command. This ensures that no files will be in use.
7. Move the /usr/omni/config (on HP-UX 9.0x) or /etc/opt/omni/
(on HP-UX 10.x) directory and recursively copy the files restored to the
temporary directory to these directories. This recreates the previous
configuration.
8. Start OmniBack II processes with the omnisv.sh -start command.
9. Start the graphical user interface and restore all other files from your
backup.
Your Cell Server should be successfully recovered.
Example of mission critical system recovery
This section provides an example how you can prepare for a crash and
recover a mission critical system, such as an OmniBack II Cell Server.
This example automates the recovery process.
This example uses unsupported scripts provided by OmniBack II.
❒ An operating system and all mission critical applications (OmniBack II
Cell Server) are installed on the same disk (same volume group).
❒ The OmniBack II software, including the Cell Server, Disk Agents, and
Media Agents isinstalled on the system.
❒ You will need a new disk of the same size and with the same partition
configuration as the original disk.
❒ You will need a standalone DAT connected directly to the system.
Preparing for a crash
1. Make sure that all the filesystems or volumes of the critical disk are inone
volume group and on the same physical disk, for example, vg00.
7-23
OmniBack II Advanced Tasks and Concepts
Recovery after a Crash
2. Print out the information about the volume group configuration. For each
logical volume you need to know where it ismounted (or the swap type)
and its size. You must save this information!
3. Configure the standalone DAT in OmniBack II. See “Configuring and
Using Logical Devices” on page 3-25 for instructions.
4. Initialize and label medium you will use for backup, for example
CRITSYS_OB2. See “Media Management Tasks” on page 3-37 for
instructions.
5. Create a datalist with all filesystems on the critical disk. Set the device
concurrency to 1 for a faster restore. Exclude the /var/opt/omni file
and /dev/vg00 (a device file for adisk with critical applications). These
files must be exluded so that they are not overwritten during restore. See
“Understanding Datalists” on page 4-3 for more information.
Print out the datalist and save the printout.
6. Run a backup of this datalist and remember the sessionID.
7. Run the /opt/omni/sbin/utilns/omnipfr.ksh sessionID
script. This saves all OmniBack II relevant information to the
/OB2RECOVERY directory.
8. Back up this directory using the tar command to an empty DAT medium
and label it, for example, OB2RECOVERY_tar. Do NOT run this on the
DAT used in previous steps.
cd /OB2RECOVERY
tar cvfh device
This creates the recovery tape with OmniBack II software needed for
restore.
At the end of these steps you have the following:
❒ A printout of the volume group information.
❒ A printout of the datalist used for backup of the system.
❒ A OmniBack II DAT tape with a backup of the system.
❒ A tar DAT tape with all OmniBack II software needed for recovery.
You should save all this in a safe place.
7-24
OmniBack II Advanced Tasks and Concepts
Recovery after a Crash
Recovering from a crash
1. Replace a crashed disk with a new disk of the same size.
2. Install the HP-UX operating system from your system installation tape
and use the printout of the volume groups to reconfigure the logical
volumes.
3. Insert the tar formatted recovery DAT tape (OB2RECOVERY_tar) in the
device and run:
cd /
tar xvf device
This installs back all OmniBack II software needed for recovery of the
system.
4. Insert the OmniBack II formatted DAT with a backup of the whole system
(CRITSYS_OB2) and run:
/omnirecover.ksh device
This restores all files on their places. If the file is busy, it is first renamed
to #filename and then the original file is restored.
5. Reboot the system.
All files that were backed up are now restored. An exception are the
excluded files. Some files, however, still differ: the files that have been
created during the installation of the operating system but were not on
the recovery tape are removed.
6. Restore OmniBack II database. See “Restoring the Database” on
page 8-14 for instructions.
Now you can restore all other files from previous OmniBack II backups.
7-25
OmniBack II Advanced Tasks and Concepts
Moving a Cell Server
Moving a Cell Server
Sometimes you may need to move the Cell Server to another system. This
section explains how to move a Cell Server if it is running properly. In case
of a system crash, refer to the “Recovery after a Crash” on page 7-20 section
first.
Make sure that you prepare carefully for this task.
1. Decide which system will be a new Cell Server and install it as a Cell
Server - see the HP OpenView OmniBack II Installation Guide for
instructions. Note that you must request to move your licenses to the new
Cell Server.
2. Make a backup of your cell.
3. Make a backup of your database - see “Backing Up the Database” on
page 8-12 for instructions.
Moving a Cell Server
On the old Cell Server, perform the following:
1. No backup or restore session should be running. Run the omnistat
command to check this.
2. Export all hosts from the cell. See “Exporting hosts from a cell” on
page 2-6 for instructions.
3. Exit the graphical user interface.
4. Run the omnisv.sh -stop command to stop all OmniBack II processes.
On the new Cell Server (which must be already installed), perform the
following:
1. Move the /usr/omni/config (on HP-UX 9.0x) or /etc/opt/omni/
(on HP-UX 10.x) directory.
7-26
OmniBack II Advanced Tasks and Concepts
Moving a Cell Server
2. Recursively copy the /usr/omni/config (on HP-UX 9.0x) or
/etc/opt/omni/ (on HP-UX 10.x) directory from the old Cell Server
to the new one.
3. Edit appearances of the old Cell Server name in the following new files.
On HP-UX 9.0x:
/usr/omni/config/cell/cell_server
/usr/omni/config/cell/cell_info
/usr/omni/config/users/UserList
On HP-UX 10.x:
/etc/opt/omni/cell/cell_server
/etc/opt/omni/cell/cell_info
/etc/opt/omni/users/UserList
You can use the following script to do that for you:
/usr/omni/bin/install/omnichhost.ksh [-preview]
old_name new_name
(on HP-UX 9.0x)
/opt/omni/sbin/install/omnichhost.ksh [-preview]
old_name new_name
(on HP-UX 10.x) to change
4. Run the omnicc command to update the configuration.
5. Run the graphical user interface and import all hosts. See “Importing
hosts to the cell” on page 2-5 for instructions.
6. Restore or copy the database to the new Cell Server.
7. Check if users are still correctly configured for the new cell.
8. Check if logical devices are still correctly configured for the new cell.
9. If you have removed the old Cell Server from your cell, also remove all
its appearances from the configured datalists.
7-27
OmniBack II Advanced Tasks and Concepts
MC/ServiceGuard and OmniBack II
MC/ServiceGuard and OmniBack II
HP OpenView OmniBack II supports the MC/ServiceGuard software
product provided for effective recovery of a particularly valuable system in
case of a system crash.
This section provides an explanation of supported functionality, limitations,
and instructions for configuration.
NOTE:
Instructions here are given as general guidelines only.
We recommend that you first get familiar with MC/ServiceGuard by creating and
testing the example applications in the MC/ServiceGuard manuals.
Refer to the following manuals for more information:
•
Managing MC/ServiceGuard for more information on MC/ServiceGuard.
•
HP OpenView OmniBack II Installation Guide for more information on
how to install OmniBack II.
•
HP OpenView OmniBack II Software Release Notes for last minute
information on the current OmniBack II release.
Functionality and Behavior
MC/ServiceGuard is supported on HP-UX 10.x systems only.
For OmniBack II, the really critical component is the Cell Server.
MC/ServiceGuard integration provides a way to have an additional Cell
Server, called a Secondary Cell Server, which can be used instead of the
Primary Cell Server. In order to do this, the Cell Server software is viewed
as a specific application that can be migrated to the Secondary Cell Server in
case of a problem with Primary Cell Server.
7-28
OmniBack II Advanced Tasks and Concepts
MC/ServiceGuard and OmniBack II
When MC/ServiceGuard detects problems with the Primary Cell Server, it
migrates the whole Cell Server Package to the Secondary Cell Server.
Therefore, both Cell Servers must have fully installed and configured
versions of OmniBack II. Note, however, that the Secondary Cell Server
cannot be used as a Cell Server of some other cell at the same time.
If a failure of the Primary Cell Server occurs while a backup is running, the
backup will fail for the currently backed up system and all others that were
scheduled for backup after this system. However, a complete OmniBack II
operation is now running on the Secondary Cell Server where you can start
an interactive backup of failed systems.
Prerequisites for Configuration
Before you start configuring OmniBack II with MC/ServiceGuard, check the
following:
❒ Decide which systems are going to be the Primary Cell Server and the
Secondary Cell Server. See HP OpenView OmniBack II Installation
Guide for more information how to choose an appropriate system.
❒ Systems chosen to be the Primary Cell Server and the Secondary Cell
Server must have MC/ServiceGuard installed and must be configured as
cluster members. See Managing MC/ServiceGuard for instructions.
❒ Systems chosen to be the Primary Cell Server and the Secondary Cell
Server must have HP OpenView OmniBack II Cell Server software
installed. See HP OpenView OmniBack II Installation Guide for
instructions. Note that a permanent license must be installed on the
Primary Cell Server.
Configuring Primary and Secondary Cell Servers
This section provides general guidelines only. For more detailed
instructions, refer to the Managing MC/ServiceGuard manual.
7-29
OmniBack II Advanced Tasks and Concepts
MC/ServiceGuard and OmniBack II
On the Primary Cell Server:
1. Set up the /dev/vg_ob2cs volume group on a shared disk accessible to
both Cell Servers.
2. Create the logical volume /dev/vg_ob2cs/lv_ob2cs for that group.
3. Mount the logical volume to the /omni_shared directory.
4. Modify the /etc/opt/omni/cell/sg.conf template file. You must
change the SHARED_DISK_ROOT variable if you have configured some
other shared directory and the CS_SERVICE_HOSTNAME variable. This
variable tells OmniBack II the name of the Primary Cell Server. The
default name is ob2cs.net.domain. You must change the
net.domain to match your network.
5. Run /opt/omni/sbin/install/omniforsg.ksh -primary to
configure the Primary Cell Server.
6. Unmount the /omni_shared directory.
7. Deactivate the /dev/vg_ob2cs volume group.
On the Secondary Cell Server:
1. Import the /dev/vg_ob2cs volume group.
2. Mount the logical volume to the /omni_shared directory.
3. Run /opt/omni/sbin/install/omniforsg.ksh -secondary to
configure the Secondary Cell Server.
4. Unmount the /omni_shared directory.
5. Deactivate the /dev/vg_ob2cs volume group.
Configuring of the Cell Server package
This section provides general guidelines only. For more detailed
instructions, refer to the Managing MC/ServiceGuard manual.
1. Generate, modify, and distribute the package control script. Make sure
7-30
OmniBack II Advanced Tasks and Concepts
MC/ServiceGuard and OmniBack II
that you provide the following information:
•
/dev/vg_ob2cs/lv_ob2cs must be mounted as /omni_shared
•
The service command must be
/etc/opt/omni/sg/csfailover.ksh start
•
The IP address must be the address associated with the relocatable
hostname (ob2cs.net.domain)
2. Generate, modify, and distribute the cluster and package configuration
scripts.
Starting the Cell Server
This section provides general guidelines only. For more detailed
instructions, refer to the Managing MC/ServiceGuard manual.
1. Halt the cluster.
2. Modify and distribute the cluster configuration file. If you are already
using MC/ServiceGuard, this file already exists. Otherwise, you must
create it.
3. Start the cluster.
4. Hand over control of the volume group /dev/vg_ob2cs to
MC/ServiceGuard using the vgchange command. See man pages for
details.
5. Make sure that switching of the OB2SG package is enabled. See man page
for the cmmodpkg command for details.
6. When the Primary Cell Server is restarted, proceed with the installation of
OmniBack II client systems.
7-31
OmniBack II Advanced Tasks and Concepts
MC/ServiceGuard and OmniBack II
Backing up a Secondary Cell Server system
If you want to back up the Secondary Cell Server or run the OmniBack II
GUI on it, you must run the following command on the Secondary Cell
Server:
mkdir -p /omni_shared/etc_opt_omni/cell
Now you can import the host to your OmniBack II cell configuration,
distribute the Disk Agent, and configure the system as any other client.
7-32
OmniBack II Advanced Tasks and Concepts
Notification and Reporting Scripts
Notification and Reporting Scripts
Administrators of larger environments often need to generate statistical or
functional reports on backup behavior such as:
•
Which devices are used for which datalist?
•
Are Post- exec commands properly configured in all scheduled datalists?
•
What filesystems are not configured for backup?
•
What is the status of last night’s backup?
Example scripts are provided which can be modified to suit specific needs:
Figure 7-1
omninotify.ksh
provides reports suitable for email or WWW notification.
omnirpt.ksh
reports OmniBack II configuration.
The HTML output window of the omninotify.ksh script
7-33
OmniBack II Advanced Tasks and Concepts
Notification and Reporting Scripts
The omninotify.ksh script
The /usr/omni/bin/omninotify.ksh (on HP-UX 9.0x) or
/opt/omni/bin/omninotify.ksh (on HP-UX 10.x) script is intended to
be used to report the status of backup sessions at regular time intervals. For
example, you can schedule the script (using cron) to send you email about
the status of a backup session in the night.
The script does not provide reports for barlists.
-wo t [n]
Generates a report on backup sessions run in
the last t hours. If may add additional number n
that specifies the duration.
This option can be combined with the html
option.
Example:
omninotify.ksh -wo 12
generates report on all backup sessions that
have been run in the last 12 hours.
omninotify.ksh -wo 12 5
generates a report on all backup sessions that
have been run in the time span of 5 hours
starting 12 hours ago.
-detail <sessionID>
Generates a report on a specified session
showing detailed information on the session.
Example:
omninotify.ksh -detail 1996/02/21-5
generates a detailed report on the session with
sessionID 1996/02/21-5.
-html <html_dir> -wo <t>
[<n>]
Generates a wo.html HTML file in the
<html.dir> directory with a table as generated
by the -wo option. Each listed datalist is a link
7-34
OmniBack II Advanced Tasks and Concepts
Notification and Reporting Scripts
to an ascii file with a detailed report for the
appropriate session as is generated by
the -detail option.
Example:
omninotify.ksh -html /tmp -wo 12
generates a report on backup sessions that have
been run in the last 12 hours. The report is
generated in the HTML format in the
/tmp/wo.html file.
See the omninotify.ksh man page for a description of each displayed
field.
The omnirpt.ksh script
The /usr/omni/bin/omnirpt.ksh (on HP-UX 9.0x) or
/opt/omni/bin/omnirpt.ksh (on HP-UX 10.x) script provides reports
on OmniBack II configuration, such as datalists, logical devices, media used
with a backup, scheduling, and so forth.
The script does not provide reports for barlists.
By default, a report is generated for scheduled datalists only. You can use the
-all option to report on all datalists, or the -subset option with a list of
datalists names to report on specific datalists only. This applies to all options
below.
-fs_noconf
Checks if filesystems mounted on any system in
the cell are configured in selected datalists
(scheduled by default.) A list of non-configured
filesystems is generated. Note that this
command can be time consuming, depending
on your cell configuration.
-dl_dev
Outputs a list of selected datalists with a list of
logical devices configured for each datalist.
7-35
OmniBack II Advanced Tasks and Concepts
Notification and Reporting Scripts
-dev_dl
Outputs a list of the logical devices configured
in the cell and a list of selected datalists where
logical devices are configured.
-info_dl
Outputs a list of selected datalists with their
owners, and the Pre- and Post- exec commands
specified for each datalist.
-sched_dl
Outputs a list of selected datalists with schedule
information.
-trees_dl
Outputs a list of selected datalists with
scheduling information, logical devices
configured for the datalist, the media pool
configured for devices, and the filesystem being
backed up with each datalist.
-obj_nobackup
Outputs a list of selected datalists with a list of
objects without protected full backup.
Displayed are the object type, the object name,
and the object description.
Examples:
1. To get a list of user directories not yet configured in any datalist in the cell
for either HP-UX 9.0x or HP-UX 10.x systems:
omnirpt.ksh -all -fs_noconf | grep -E "home | users"
To get a list of filesystems not yet configured in scheduled datalists in the
cell (these filesystems will not be backed up with regular scheduled
backup):
omnirpt.ksh -fs_noconf
2. To get a list of all logical devices used in a datalist named local:
omnirpt.ksh -subset local -dl_dev
See the omnirpt.ksh man page for a description of each displayed field.
7-36
OmniBack II Advanced Tasks and Concepts
Customizing the Compression Algorithm
Customizing the Compression Algorithm
To reduce the number of media needed to perform a backup, you can enable
the compress option using the backup options window. The compression
used by OmniBack II is the Lempel-Ziv 4.3 compression algorithm. This is
the same compression algorithm used by the Unix compression utility and
provides reasonable compression and speed for most data types. However,
you may choose to implement another compression algorithm, customized
for your type of data. Customizing the compression algorithm may provide
better performance for the backup and restore of specific types of data (for
example, graphics-oriented applications.)
To customize the compression algorithm, you can write your own
compression code module, compile the module, and link it to the standard
HP OmniBack II compression library. This library is called libdc.sl and
is located in the /usr/omni/lib (on HP-UX 9.0x) or /opt/omni/lib
(on HP-UX 10.x) directory. The following section describes how to create
your own compression module. For detailed information about developing
Unix code, consult the HP-UX Programming Manual.
CAUTION:
If you change the compression algorithm, OmniBack II will no longer be able to
decompress data from backups performed with the previous algorithm. If this occurs,
OmniBack II will restore the data in compressed form. Therefore, it is recommended
that you only change the algorithm after careful planning. Frequent changes to the
compression algorithm increase the risk of data loss.
CAUTION:
To restore a backup created using a different compression algorithm, you must first
restore the version of the compression algorithm that was used to create the backup
set you want to restore.
7-37
OmniBack II Advanced Tasks and Concepts
Customizing the Compression Algorithm
Creating a compression code module
This section provides an overview of the process for developing your own
code module for compression. For more detailed information, consult the
HP-UX Programming Manual.
Creating and installing a new compression code module involves the
following tasks:
1. rename the original shared library (libdc.sl)
2. write the new code module
3. compile the new code module
4. link the new code module to the original shared library
5. rename the new shared library to the original shared library (libdc.sl)
NOTE:
Since the compression shared library is accessed by the backup and restore Disk
Agents, the shared libraries must be replaced on all hosts where the Disk Agents run,
not just on the Cell Server host.
The data compression type identifiers used in the compression module are
shown here:
No data compression
DC_LZ43 0x0b0b0b0b Lempel-Ziv 4.3
DC_NONE 0
The following list briefly explains each function used and its purpose.
DcInit
DcCompress
DcDecompress
DcExit
initializes the compression module and returns the
compression type
the compression function
the stream decompression function
performs post-processing cleanup
The following section shows the data compression modules, with an
explanation following each function.
Module Definition - Data Compression Type Identifiers
7-38
OmniBack II Advanced Tasks and Concepts
Customizing the Compression Algorithm
#if !defined(UCHAR)
#define UCHAR
unsigned char
#endif
#if !defined(ULONG)
#define ULONG
unsigned long
#endif
#define
DC_NONE
0
/* No data compression */
#define
DC_LZ43
0x0b0b0b0b /* Lempel-Ziv 4.3 */
typedef int (*dcCallback)(UCHAR *, ULONG);
MODULE
Data Compression (Dc)
FUNCTION
extern ULONG
DcInit (
#if PROTOTYPES
void
#endif
);
ARGUMENTS
void
RETURNS
DC_NONE
DC_LZ43
DC_xxxx
Initialization failed, compression disabled
Initialization successful, Lempel-Ziv compression
a customer-created & implemented compression type,
where xxx represents a customer-specified name
DESCRIPTION
This function is called by the Disk Agent’s initialization procedure at
startup. It should initialize the Data Compression module and indicate the
compression type.
REMARKS
Dynamic switching between different compression algorithms depending on
the type of input data is not available.
MODULE
7-39
OmniBack II Advanced Tasks and Concepts
Customizing the Compression Algorithm
Data Compression (Dc)
FUNCTION
extern int
DcCompress (
#if PROTOTYPES
dcCallback
dcCallback
#endif
);
readCallback,
writeCallback
ARGUMENTS
readCallback
writeCallback
A pointer to the data input callback
A pointer to the data output callback
RETURNS
0
1
Error
OK
DESCRIPTION
This is the actual stream compression function. Input is received from the
readCallback function and output is written using the writeCallback
function. Note the following guidelines when modifying this function:
•
The compression module cannot detect the type of input and output
media. Therefore, the input should be treated as a read-only stream and
the output should be treated as a write-only stream.
•
The DcCompress function is called once for each file to be compressed.
It reads the input until EOF (the input handler returns a value of zero) and
makes sure that all the intermediate output buffers are flushed before it
sends a return value.
•
Each call to the readCallback is directly translated into a read() system
call. Therefore, for better performance, the compression engine should
read its input in large chunks and store them in an internal input data
buffer.
•
The readCallback function returns the number of bytes read. This number
can, under certain circumstances, be less than the number of bytes actually
requested. The compression engine should not treat this as an error but
simply issue a second read (this is similar to BSD sockets, reads may
succeed partially). The readCallback returns a value of zero when an error
7-40
OmniBack II Advanced Tasks and Concepts
Customizing the Compression Algorithm
occurs or at EOF.
•
The writeCallback function always sends a return value. If an error
occurs, the writeCallback return value is zero.
The readCallback return values are:
>0
0
Number of bytes read
Error or EOF
The writeCallback return values are:
>0
0
number of bytes written
Error
REMARKS
Never call the readCallback to read or write zero bytes, as the return value
will be zero which means ’Error or EOF’.
MODULE
Data Compression (Dc)
FUNCTION
extern int
DcDecompress (
#if PROTOTYPES
dcCallback
dcCallback
#endif
);
readCallback,
writeCallback
ARGUMENTS
readCallback
writeCallback
A pointer to the data input callback
A pointer to the data output callback
The return values are:
0
1
Error
OK
DESCRIPTION
7-41
OmniBack II Advanced Tasks and Concepts
Customizing the Compression Algorithm
This is the actual stream decompression function. It receives its input data
using the readCallback data input function and writes its output using the
writeCallback output function. Note the following guidelines when
modifying this function:
•
The nature of the input and output media is unknown to the
decompression module. Therefore, the input should be treated as a readonly stream and the output should be treated as a write-only stream.
•
The DcDecompress function is called once for each file to be
decompressed. It reads its input until EOF (the input handler returns zero)
and checks that all intermediate output buffers are flushed before it sends
a return value.
•
Both input and output can be done in byte chunks of arbitrary length. The
readCallback function returns data from an internal OmniBack II data
buffer. However, for the best performance, it is recommended that output
be sent in larger chunks, since calls to the writeCallback function can be
directly translated into write() system calls.
•
The readCallback function can partially succeed. A return value of zero
denotes either an error or the end of the current file to be decompressed.
The writeCallback function always succeeds completely and never
reports any errors back.
The readCallback return values are:
>0
0
Number of bytes read
Error or EOF
The writeCallback return values are:
>0
0
Number of bytes written
Error or EOF
REMARKS
Never instruct the readCallback function to read zero bytes, because this
results in a return value of zero, which means ’Error or EOF’. For the same
reason, never attempt to write zero bytes.
MODULE
Data Compression (Dc)
7-42
OmniBack II Advanced Tasks and Concepts
Customizing the Compression Algorithm
FUNCTION
extern int
DcExit (
#if PROTOTYPES
void
#endif
);
ARGUMENTS
void
RETURNS
0
1
Error
OK
DESCRIPTION
This function allows the DC module to perform post-processing cleanup
actions. Its return value is ignored.
7-43
OmniBack II Advanced Tasks and Concepts
Customizing the Encryption Algorithm
Customizing the Encryption Algorithm
To increase the security of your backup media, you can enable the encrypt
option using the backup options window. This backup option writes data to
the media in an encrypted form. Media encrypted using the HP OmniBack II
encryption option can only be read by an OmniBack II Media Agent using
the same encryption key.
To customize the encryption algorithm, you can write your own encryption
code module, compile the module, and link it to the standard HP OmniBack
II encryption library. This library is called libde.sl and is located in the
/usr/omni/lib (on HP-UX 9.0x) or /opt/omni/lib (on HP-UX 10.x)
directory. The following section describes how to create your own
encryption module. For detailed information about developing Unix code,
please consult the HP-UX Programming Manual.
CAUTION:
If you change the encryption algorithm, OmniBack will no longer be able to
deencrypt data from backups performed with the previous algorithm. If this occurs,
OmniBack II will restore the data in encrypted form. Therefore, it is recommended
that you only change the algorithm after careful planning. Frequent changes to the
encryption algorithm increase the risk of data loss!
CAUTION:
To restore a backup created using a different encryption algorithm, you must first
restore the version of the encryption algorithm that was used to create the backup set
you want to restore.
Creating an encryption code module
This section provides an overview of the process for developing your own
code module for encryption. For more detailed information, please consult
the HP-UX Programming Manual.
Creating and installing a new encryption code module involves the
following tasks:
7-44
OmniBack II Advanced Tasks and Concepts
Customizing the Encryption Algorithm
1. rename the original shared library (libde.sl)
2. write the new code module
3. compile the new code module
4. link the new code module to the original shared library
5. rename the new shared library to the name of the original shared library
(libde.sl)
NOTE:
Since the encryption shared library is accessed by the backup and restore Disk
Agents, the shared libraries must be replaced on all hosts where the Disk Agents run,
not just on the Cell Server host.
The constants used in the encryption module are shown here:
DE_NONE 0
no encryption
DE_XOR 1
simple XOR encryption
The following list briefly explains each function used and its purpose.
define
identifies the encryption type
DeInit
initializes the encryption module
DeEncrypt
performs the encryption
DeDecrypt
performs the deencryption
DeExit
exits the encryption module
The encryption module is shown with an explanation of the functions in the
module. Module Defines - Unique Data Encryption Type Identifiers
#if !defined(UCHAR)
#define UCHAR unsigned char
#endif
#if !defined(ULONG)
#define ULONG unsigned long
#endif
#define DE_NONE 0 /* No data encryption */
#define DE_XOR 1 /* Simple XOR encryption */
#define DE_DES 2 /* DES encryption (not impl.) */
7-45
OmniBack II Advanced Tasks and Concepts
Customizing the Encryption Algorithm
MODULE
Data Encryption (De)
FUNCTION
extern ULONG
DeInit (
#if PROTOTYPES
void
#endif
);
ARGUMENTS
void
RETURNS
DE_NONE
Initialization failed, encryption disabled
DE_XOR
Initialization successful, xor encryption
DE_xxxx
Customer-created & implemented encryption type, where
nnnn represents a customer-supplied name
DESCRIPTION
This function is called by the Disk Agent’s initialization procedure at
startup. It initializes the DE module and retrieves the encryption key from a
file or key server.
REMARKS
The distribution of the encryption key through the network is a complex
issue and not within the scope of a backup/restore application. This function
can read a key from a predefined file or contact a key distribution server (via
RPC) to obtain a key.
MODULE
Data Encryption (De)
FUNCTION
extern int
DeEncrypt (
#if PROTOTYPES
UCHAR *data,
ULONG size
#endif
);
7-46
OmniBack II Advanced Tasks and Concepts
Customizing the Encryption Algorithm
ARGUMENTS
data
size
pointer to plain text buffer
size of plain text buffer
RETURNS
0
1
Error
OK
DESCRIPTION
The encryption of the plain text should be done in-place. OmniBack II
expects the size of the data buffer to be maintained and the contents to be
overwritten with the encrypted text.
NOTE: The encryption algorithm must generate encrypted text of the same
length as the original text.
MODULE
Data Encryption (De)
FUNCTION
extern int
DeDecrypt (
#if PROTOTYPES
UCHAR *data,
ULONG size
#endif
);
ARGUMENTS
data
pointer to crypt text buffer
size
size of crypt text buffer
RETURNS
0
1
Error
OK
DESCRIPTION
The decryption of the crypt text should be done in-place. OmniBack II
expects that the size of the data buffer will remain the same and that the
contents will be overwritten with the decrypted text.
7-47
OmniBack II Advanced Tasks and Concepts
Customizing the Encryption Algorithm
MODULE
Data Encryption (De)
FUNCTION
extern int
DeExit (
#if PROTOTYPES
void
#endif
);
ARGUMENTS
void
RETURNS
0
1
Error
OK
DESCRIPTION
This function is called to allow the DE module to perform a post-processing
cleanup. Its return value is ignored.
7-48
8
Maintaining the OmniBack II Database
8-1
Maintaining the OmniBack II Database
In This Chapter
This chapter describes how to maintain and back up the OmniBack II
database. This database is used to record the following information:
•
•
•
•
•
•
•
•
configured devices, media, and pools
usage and condition of the media in the pools
date and time backups are started
type of backups
directories, files and disks that were backed up
file positions on the media
messages resulting from a backup or restore session
expiration date for the data contained on the backup media
Maintenance is needed to control the size of the database, and periodic
backups of the database ensure its integrity.
Managing the database used to store this information consists of three tasks:
•
•
•
Managing the growth of the database
Backing up the database for recoverability
Maintaining the integrity of the database
Each of these tasks is explained in detail in this chapter.
8-2
Maintaining the OmniBack II Database
Installing and Configuring the Database
Installing and Configuring the Database
The OmniBack II database is installed to the /usr/omni/config/db (on
HP-UX 9.0x) or /var/opt/omni/db (on HP-UX 10.x) directory on a local
disk.
NOTE:
Database must be located on the local disk and not on NFS mounted disks
The database consists of several types of files:
omni.dbd
data dictionary
*.dat
data files
*.key
index [key] files
*.ext
string extensions - data
8-3
Maintaining the OmniBack II Database
Managing the Growth of the Database
Managing the Growth of the Database
The ability to locate and quickly recover files is an important part of HP
OmniBack II. Approximately 80 to 85% of the database is occupied by the
detail catalog. This detail catalog is a list of all the backed up files and
directories. The fvers.dat is the biggest file of the detail catalog.
OmniBack II uses the size of the fvers.dat file to calculate the size of the
database.
A general estimate for the size of the database is approximately 1% of the
data being backed up. Thus, if the data is 100GB, the database will occupy
approximately 1GB.
NOTE:
Make sure that you have enough free space on a disk for the database. If the database
runs out of disk space during backup, the database may become inconsistent.
Projecting the growth of the database
The growth and size of the OmniBack II database are determined by the
following factors:
•
the number of logged messages
•
the amount of details on backed up files and directories stored in the
database
•
the number of backup and restore sessions
Changing the number of logged messages
The level of detail recorded in the database relates to the Report Level and
the file logging of the backup options. If you set the Report Level to
warning, all screen messages at the warning level and higher are recorded
in the database. The number of messages recorded will be higher than if you
set the Report Level to critical.
8-4
Maintaining the OmniBack II Database
Managing the Growth of the Database
See “Selecting the filesystem backup options” on page 4-10 and “Selecting
the rawdisk backup options” on page 4-13 for instructions.
Changing the amount of details on files stored in the database
OmniBack II provides three levels of the amount of details on files and
directories stored in the database.
Log All
This is the default option. All backed up files are logged
to the database. This complete information allows you to
search for backed up files and allows OmniBack II to fast
position on the tape when restoring a specific file.
However, this information may take a lot of space if there
are many files.
Log Directories Details on directories only are stored in the database. This
disables the search feature during restore and you will
only be able to browse directories. However, OmniBack
II still performs fast positioning because a file is located
on the tape near the directory where it actually resides.
This option is suitable for filesystems with many autogenerated files, such as news and mail systems.
No Log
No details on files or directories are logged in the
database. You will not be able to search and browse files
and directories. The restore will take longer, because
OmniBack II cannot fast position on the tape but will read
from the start of the backup.
See “Selecting the filesystem backup options” on page 4-10 and “Selecting
the rawdisk backup options” on page 4-13 for instructions.
Changing the number of backup and restore sessions
The other factor in estimating database growth is the number of backup and
restore sessions. A schedule with frequent full backups will result in a faster
growth of the database.
8-5
Maintaining the OmniBack II Database
Managing the Growth of the Database
Limiting the growth of the database
The size of the database is limited by two factors: the limit you set in the
global options file (see “OmniBack II Global Options File” on page 7-11)
and the amount of free space available on the disk where the database is
located.
NOTE:
When you set the limit for the database, you are actually setting the size limit for the
fvers.dat file. This file is part of the detail catalog and occupies about 65-70% of
the database.
Setting a Limit
You can set a limit for the maximum database size by inserting the following
line in the global option file:
DBLIMIT=size
The size variable represents the maximum number of bytes available for the
database. The default setting is 1.96 GB to prevent exceeding UNIX file size
limit (2 GB.)
For example, you would enter 15000000 to represent 15 MB.
It is recommended that you set the limit to 70% of the estimated amount of
space needed for the database. For example, if you are backing up 15 GB of
data, the limit is calculated as follows:
15 GB x .01 = 0.15
0.15 GB x .70 = 100 MB
In this example, you would enter 100000000 (100 MB) as the value for
DBLIMIT.
When this limit is reached, the following error message is displayed:
Size limit of the database reached. Catalogs will not be written to
the database for the rest of this session.
8-6
Maintaining the OmniBack II Database
Managing the Growth of the Database
When this occurs, the Session Manager stops writing the name of each file to
the database. Note that session and object information continues to be
written to the database and that all currently running sessions still complete
successfully.
Since session and object information is still written to the database, the
database can continue to grow, but at a slower pace. When there is no free
space available on the disk where the database is located, all currently
running sessions terminate. This can result in a corrupted database.
It is recommended that when the user-defined limit is reached, you either
increase the limit or purge obsolete data from the database.
Purging obsolete data from the database
Purging obsolete information from the database is an important part of
database maintenance. It removes unneeded information from the database
and frees space for new information. Note that by purging the database you
do not actually shrink the size of the file - you only make space for new
information. It is good to do this maintenance on a regular basis.
All these actions must be executed on the Cell Server where the database is
located.
NOTE:
These commands can only be used when no backup or restore session is running. This
is because you cannot write to the database at the same time that you are purging data
from it. You must exit the GUI before purging data.
HP OmniBack II provides the DBPurgeBeforeBackup and
DBPurgeFilesBeforeBackup options to allow for automatic purge of
unneeded data from the database before a database backup. By default, these
options are not set. See “OmniBack II Global Options File” on page 7-11 for
more information how to set these options.
8-7
Maintaining the OmniBack II Database
Managing the Growth of the Database
To purge obsolete restore data
You can purge the database by deleting data related to restore sessions. The
following command purges information relating to restore sessions as well
as obsolete (overwritten) backup sessions and sessions without any media
such as failed or aborted sessions.
OmniBack II purges all information relating to restore sessions older than
the number of days that you specify. The command is shown here:
omnidbcheck -purge days
To purge the detail catalog from the database
OmniBack II records the names of each file being backed up during each
session in the detail catalog. You can create space for new data by purging
this detail catalog from the database. You can purge the detail catalog for a
backed up object in one specific backup session. The information that
remains in the database consists of the object name, medium ID, and the date
and time of the backup, which is sufficient to recover all backed up data, but
no browsing of files is available. The restore of single files will be slower
since OmniBack II has to search from the beginning of the media for the file.
You can mark the detail catalog for purging using the omnidb -strip
command with various options. See the omnidb manpage for a complete
description of the command options.
Example:
omnidb -strip
marks the details catalog of all objects with expired protection.
If you combine this command with -session sessionID, all objects
(filesystems) backed up in this session are marked.
If you combine this command with -session sessionID and filesystem filesystemID, the specified filesystem in the specified
session is marked in the details catalog.
8-8
Maintaining the OmniBack II Database
Managing the Growth of the Database
If you combine this command with days -filesystem
filesystemID, all versions of the specified filesystem older than days days
are marked in the details catalog.
NOTE:
These commands do not actually delete data from the database, but just marks the
detail catalog for deleting. To actually purge the database, you must run the
omnidbcheck -purge command afterwards.
TIP:
We recommend running the omnidbcheck
-purge command
regularly. You may want to configure an automatic purge of the database
before the database backup by setting the DBPurgeBeforeBackup
variable in the global options file.
To purge an unneeded session from the database
HP OmniBack II keeps track of how long data on a medium is protected (see
“Using Backup Options” on page 4-19 for more information). When the
protection expires, the session will be overwritten by newer backups.
OmniBack II provides a way to delete an unwanted session from the
database.
1. From the Monitor window choose View → Previous Sessions.
2. Select the session you want to delete and choose Actions → Purge
Session.
The session disapears from the list.
The following happens:
•
Protection is removed from the session. This means that the session will
be overwritten during future backups.
•
All detail information (backed up files and messages) about the session is
removed from the database.
•
The session itself remains on the media until it is overwritten by the next
backup. Therefore, you can still see or restore the session using list from
media.
8-9
Maintaining the OmniBack II Database
Managing the Growth of the Database
Shrinking the size of the database file
By using the above procedures you do not really shrink the size of the
database, you just remove obsolete or detail information and create more
space for new. If you really must reduce the size of a database file, the only
possible way is to copy the database to an ASCII file and then read it back.
To ensure the integrity of the database, make sure that backups, restores, or
GUIs are not running before copying the database.
Is it feasible to shrink the database?
You can use the omnidbcheck -info command to see how many records
are actually used and how many are occupied by data. If the numbers vary
considerably, it is feasible to shrink the database.
Copying the database to an ASCII file
Before copying the database to an ASCII file we recommend running the
omnidb -purge command.
Use the following command to copy the database to an ASCII file.
omnidbcheck -writeascii directory [-no_details]
If you specify the -no_details option, the detail catalog is not copied.
NOTE:
For this operation you will need approximately 50 % more disk space as is the size
of your database.
This operation may also take considerable time, for example, writeascii and readascii
on an HP Series 800 E35 system the writeascii of the 500 Mb database may take 5
hours.
8-10
Maintaining the OmniBack II Database
Managing the Growth of the Database
Reading the database from an ASCII file
Note that by reading the database from the ASCII file you will lose any data
written to the database after it was copied to the ASCII file.
Use the following command to read the database from the ASCII file.
omnidbcheck -readascii directory [-no_details]
If you specify the -no_details option, the detail catalog information is
not read and you will lose all detail information of backed up files.
If you encounter problems with this command, you should first initialize the
database with the omnidbinit command.
8-11
Maintaining the OmniBack II Database
Backing Up the Database
Backing Up the Database
The database is an extremely important part of OmniBack II and must be
backed up on a regular basis. We recommend the following:
•
Create a separate datalist for the database. This simplifies scheduling and
restoring in case of a disk crash.
•
Schedule a database backup every night. This ensures you always have a
up-to-date backup of the database. You can set the protection to only a few
days.
•
Make the database backup on a separate medium, if possible, on a
separate device, for example, to a standalone DAT. This greatly simplifies
eventual restore, since you know precisely on which medium your
database is backed up.
•
Create and use a separate pool for backing up your database. This
simplifies the recovery of the database.
The database is in use whenever you are using the graphical user interface or
when a backup or restore session is in progress. OmniBack II provides an
automatic action to back up the database that shuts down the database, backs
it up, and then restarts it.
OmniBack II also maintains the integrity of the database by preventing you
from backing up a corrupted database. When you initiate a backup of the
database, OmniBack II first performs a database check. If the database is
corrupt, the backup fails. This prevents you from later restoring a corrupted
database.
The database backup backs up all OmniBack II configuration data, not just
the database. The following directories are backed up:
On HP-UX 9.0x systems /usr/omni/config (with
/usr/omni/config/sessions excluded.)
On HP-UX 10.x systems /etc/opt/omni and /var/opt/omni/db.
8-12
Maintaining the OmniBack II Database
Backing Up the Database
CAUTION:
Although you can include the database in a regular filesystem backup, not all the
database files will be backed up. If you try to restore from this backup, you will
corrupt the database.
Backing up the database using a datalist
The database is in use whenever you are using the graphical user interface or
when a backup or restore session is in progress. OmniBack II provides you
with an automatic action that shuts down the database, backs it up, and then
restarts the database.
Backing up the database consists of these two steps:
•
•
Adding the object (OmniBack II Database) to a datalist
Performing a backup using this datalist
Creating a datalist to back up the database
1. From the OmniBack II Backup Editor window choose Object →
Add → OmniBack II Database menu selection to add the database to a
new or existing datalist.
Running backups of the database
Once you have created the datalist for the database, you can schedule the
datalist or run an interactive backup.
You can use the omnib command with the -datalist option to perform
the database backup. See the omnib man page for details.
8-13
Maintaining the OmniBack II Database
Restoring the Database
Restoring the Database
If you have backed up the OmniBack II database using the procedure
described in “Backing Up the Database” on page 8-12, you can recover the
database after restoring it to a different directory.
Restoring the Database consists of two basic steps:
1. Restoring the database to an alternative directory
2. Recovering the database
CAUTION:
Please note the following cautions for the recovery of the database:
•
You must restore the database from a backup that was created using the
method described in the “Backing Up the Database” on page 8-12. If you
try to restore the database from a regular backup, you will corrupt the
database!
•
You must restore the database into an alternative directory because the
database is in use during any restore session. The -into option is used to
specify the alternative directory for the database. If you do not use the
-into option, you will get an error message.
•
It is recommended that you create a copy of all database files by first
shutting down all OmniBack II processes using the omnisv.sh -stop
command and then copying all database files to a temporary directory.
Graphical User Interface Method
To restore the database using this method, you use the normal procedure for
restoring single files, except that you use the Fileset Restore Options
window to specify an alternative directory for the database. To restore the
database, follow these steps:
1. From the OmniBack II Main menu, select Restore to open the
OmniBack II Restore window with a list of all the objects which have
been backed up.
2. Double click the OmniDb object to open the Database Restore
window.
8-14
Maintaining the OmniBack II Database
Restoring the Database
3. Choose Actions → Restore OmniBack II DB… to open the
Filesystem Restore window.
By default, the database is restored to the /tmp directory.
To select specific restore options, choose Database Restore Options.
4. Choose OK to open the OmniBack II Restore Monitor window
where you can start restore.
Command-Line Interface Method
You can use the -omnidb option to restore the database. The command
syntax is shown here:
omnir -omnidb Hostname:Mountpoint [label] -session sessionname
-into directory name
If you specified the label when backing up the database, you must now
specify the label when restoring the database.
The -into option is required when restoring the database.
NOTE:
Do not restore the database to the root directory.
Recovering the Database
If you used the graphical user interface method or the command-line method
to restore the database, you must copy the restored database from the
alternative directory back to the original directory. When the restore of the
database is finished, follow these steps:
1. Stop all running OmniBack II sessions. If the graphical user interface is in
use, exit the interface. This prevents access to the database.
2. Use the omnisv.sh -stop command to stop all OmniBack II deamons.
3. Copy the database files from the alternative directory to the original
directory /usr/omni/config/db (on HP-UX 9.0x) or
/var/opt/omni/db (on HP-UX 10.x.)
8-15
Maintaining the OmniBack II Database
Restoring the Database
4. Use the omnisv.sh -start command to restart the OmniBack II
daemons.
NOTE:
With the database backup, all configuration files such as datalists, barlists, scheduling
information, cell configuration, were also backed up.
You can also recreate these files by copying the complete directory structure.
8-16
Maintaining the OmniBack II Database
Moving the Database to Another System
Moving the Database to Another System
HP OmniBack II provides two ways for you to move the OmniBack II
database to another system. You can either move the database directory to
another system or copy the database to an ascii file and then move the ascii
file to the other system. This is useful if you would like to move the database
to a larger system or you change the Cell Server system. Both methods are
described below.
Stopping all OmniBack II processes
Before you can move the database, you must stop all running sessions, exit
all GUIs, and run the omnisv.sh stop command to stop all OmniBack II
daemons.
Moving the Database Directory
You can move the database to another system by moving the
/usr/omni/config/db (on HP-UX 9.0x) or the /var/opt/omni/db (on
HP-UX 10.x) directory to the new system.
Copying the Database to an ASCII file and Moving the File
Using this method, you copy the database to an ascii file and then move the
ascii file to the new system. To do this, follow these steps:
1. Export the database to an ascii file using the following command. This
creates an ascii file with the data in the directory you specified:
omnidbcheck -writeascii directory name [-no_details]
The -no_details option prevents exporting of detail catalogs.
2. Move the file to the other system using rcp, ftp, or NFS mount.
3. Import the database on the other system using the following command:
8-17
Maintaining the OmniBack II Database
Moving the Database to Another System
omnidbcheck -readascii directory name [-no_details]
The -no_details option prevents importing of the detail catalogs.
8-18
9
Using Complex Devices and Exchangers
9-1
Using Complex Devices and Exchangers
In This Chapter
Any device that has more than a single drive and single cartridge is
considered to be a complex device. For example, exchangers provide you
access to a large volume of data within a few seconds, without operator
intervention. Allowing convenient access to a library of tape and magnetooptical cartridges, these devices can offer capacities of up to several 100
gigabytes. This chapter helps you to understand how to:
•
Manage complex devices
•
Configure a SCSI-II exchanger
•
Configure a non-standard exchanger
9-2
Using Complex Devices and Exchangers
Managing Complex Devices with OmniBack II
Managing Complex Devices with OmniBack II
OmniBack II’s media management services, device support features, and the
OmniBack II configuration database work together to control the behavior of
complex devices as described in this chapter.
Complex device types
OmniBack II supports three categories of complex devices:
Device Chains
A device chain is a set of standalone devices treated as one device by
OmniBack II. All physical devices in a device chain must belong to the
same class. OmniBack II uses each device in the chain one after the other.
When the operation on the first device is finished, OmniBack II resumes
backing up or restoring from the second and then the third device until
OmniBack II reaches the last device.
Stackers
A stacker is a single device usually with only one drive. A stacker takes a
medium from a "stack" (its repository) and inserts the medium into its drive.
This exchange is always limited to ejecting the medium already in the drive
and inserting the next medium from the stack. A stacker cannot randomly
pick a medium from its repository.
Exchangers
Exchangers, also called exchangers, jukeboxes, or carousels, can randomly
pick a medium from its repository and insert the medium in the drive.
Exchangers differ in repository sizes, exchanging techniques and number of
drives. Some exchangers can even read the bar code on the media in the
repository.
9-3
Using Complex Devices and Exchangers
Managing Complex Devices with OmniBack II
How OmniBack II functions with complex devices
OmniBack II performs the following operations on complex devices:
•
•
•
•
•
•
backing up data
restoring data
initializing medium to be used for backup
verifying data off-line
list cataloging
importing
How OmniBack II operates on the different complex devices types is
explained below.
Device Chains
Backup
When OmniBack II detects the end of a medium during a backup session,
OmniBack II automatically switches to the next drive configured with the
logical device. OmniBack II will continue backing up to the next configured
drive until it reaches the end of the medium in the last drive. Then
OmniBack II issues a mount request. The user should then replace all the
media in each drive. When OmniBack II detects the mount request has been
satisfied, it starts backing up from the first drive again.
Restore
How OmniBack II operates on a device chain during restore is analogous to
how it operates during backup. OmniBack II starts scanning the first drive of
the device chain to look for the media its needs for restore. If the needed
media is not in the first drive, OmniBack II automatically starts looking for it
in the next drive and so on until OmniBack II finds the drive with the correct
medium or until it comes to the last drive of the logical device. If OmniBack
II cannot find the required medium, it issues a mount request for the last
9-4
Using Complex Devices and Exchangers
Managing Complex Devices with OmniBack II
drive. If OmniBack II finds the required medium, but still needs another
medium to complete the restore, OmniBack II scans the device chain from
the current position to the end of the device chain for the needed medium.
Initialize
OmniBack II scans the device chain from the first drive in the chain to the
last for the medium which is to be initialized.
Verification
OmniBack II scans the device chain from the first drive in the chain to the
last for the medium to be verified. If OmniBack II does not find the medium,
it issues a mount request and starts scanning from the first drive again.
Stackers
Backup
OmniBack II performs backups to a stacker in the same manner as for device
chains. The only difference is that instead of changing the drive currently in
use, OmniBack II picks the next medium from the repository. If no medium
is available OmniBack II issues a mount request.
Restore
OmniBack II performs restores from stackers in the same manner as for
device chains. The media in the drive is exchanged until the desired medium
is reached. If OmniBack II needs another medium to complete the restore,
OmniBack II inserts media from the stack until the needed medium is found.
If no media is available, OmniBack II issues a mount request.
Initialize
OmniBack II initializes the medium that is at the beginning of the stack.
Verification
The specified number of media at the top of the stack is verified.
9-5
Using Complex Devices and Exchangers
Managing Complex Devices with OmniBack II
Exchangers
Backup
OmniBack II keeps a record of which media is in the repository of an
exchanger. This information is updated when a medium is entered or ejected
from the repository. OmniBack II always tries to allocate the media it will
use for backup from the contents of the repository. If OmniBack II cannot
backup using media from the repository or if OmniBack II must use a
medium from the top of the Prealloc List, OmniBack II will issue a mount
request.
If media in the repository is manually taken out (you use the Eject button)
you must rescan the repository, so OmniBack II can update its Media
Management database. For details on using the rescan function, see
“Rescanning a Device” on page 3-35.
NOTE:
We recommend to always operate the exchanger via the OmniBack II software. If you
circumvent the software and do operations on the exchanger manually, you risk
confusing the software which will result in misleading operations messages.
Restore
If the media is in the repository then it is automatically inserted in the drive
in the correct sequence and OmniBack II proceeds with the restore.
Otherwise, OmniBack II issues a mount request.
Initialization
Specify labels for all the media in the repository that is to be initialized.
Verification
Specify media labels for all the media that should be verified. OmniBack II
will only verify the specified media. If a specified media is missing,
OmniBack II issues a mount request.
9-6
Using Complex Devices and Exchangers
Configuring a SCSI-II Exchanger
Configuring a SCSI-II Exchanger
The generic tape/optical exchanger is a complex mechanical device
containing one or more data drive mechanisms, an internal magazine
(repository) of media and a mechanical picker capable of loading, and
unloading the media into the drives under software control.
The generic exchanger has a SCSI ID for each data drive and one for the
picker. For instance, an HP optical jukebox containing two optical drives has
three SCSI IDs — one for each drive and a third for the picker. One
exception to this is the HP SureStore 12000e Exchanger, which allows both
mechanisms, the data drive and the picker, to be accessed through the same
SCSI ID, but different SCSI Logical Units (lu).
Each repository slot of an exchanger holds one cartridge. OmniBack II
assigns a numerical ID to each slot within the repository starting from one.
You refer to these slots by ID numbers when you configure the logical
device.
For example, an exchanger with 6 repository slots will have the slot IDs 1, 2,
3, 4, 5, and 6. If the exchanger has multiple drives, each drive is referred to
by an ID starting with the number one too. A 4-drive exchanger, for
example, will have the drive IDs 1, 2, 3, and 4. A single drive exchanger is
referred to by the drive ID number “1.”
NOTE:
Note that the above slot and drive IDs are NOT the physical element IDs returned by
the exchanger in response to a SCSI mode sense command (ioctl). The physical
element IDs are determined by the Media Agent based on the above slot and drive
IDs and the mode sense information returned by the exchanger when a backup is run.
OmniBack II has provided you with this numbering scheme so that you do not have
to know the physical element IDs of the exchanger to complete the logical device
configuration.
If you have a multi-drive exchanger and want to use the drives in parallel,
you must configure multiple logical devices, one for each exchanger drive.
9-7
Using Complex Devices and Exchangers
Configuring a SCSI-II Exchanger
Example of an Exchanger Configuration
The following example will demonstrate how to configure an exchanger. We
will use a double-drive DAT exchanger with a repository of 50 slots and
connected to an HP9000/S700 system. The steps that follow show you how
to complete the configuration of the hypothetical logical devices, DRIVE_1
and DRIVE_2. We will configure two logical devices for this exchanger, so
we can use both drives in parallel. The first logical device will operate
through drive 1 and the second through drive 2.
The basic steps involved in configuring the exchanger are:
1. Creating an Exchanger Pool
2. Selecting the device files.
3. Configuring the logical devices.
Creating an Exchanger Pool
See “The Media Pool approach” on page 3-13 for more information on the
exchanger pool and how to create it.
Selecting the device files
NOTE:
If you’re configuring your exchanger on an HP9000/S800 system, you will have to
install the SPT driver which is available in the patch PHKL_5405 (for HP-UX 9.04
or later) or PHKL_5406 (for HP-UX 10.x.) Make sure that you follow carefully the
instructions provided with the patch. You must install the correct driver for your
devices, select the correct SCSI address, and create a device file for the device you
will use with OmniBack II.
When this driver is installed and OmniBack II performs backup or restore sessions,
we do not suggest the use of applications running ioscan, such as SAM. This could
cause the system to reboot.
9-8
Using Complex Devices and Exchangers
Configuring a SCSI-II Exchanger
1. Determine the Unix device files that correspond to the exchanger drives.
If you are configuring a tape exchanger, select the device files from the
/dev/rmt directory.
If you are using an optical exchanger, select the raw disk device files
from the /dev/rdsk directory.
The double-drive DAT exchanger in our example would use the device
files /dev/rmt/1m and /dev/rmt/2m to address the exchanger drives
(assuming the data drive SCSI IDs are 1 and 2).
2. Determine the UNIX device file that corresponds to the exchanger picker
mechanism.
Since our example uses an HP9000/S700 system, the device file
associated with the exchanger picker would be in the /dev/scsi
directory. We will use the device file /dev/scsi/0. The "0" at the end
of the device file name indicates that SCSI ID for the exchanger picker
device is 0.
NOTE:
The HP SureStore 12000e Exchanger does not have a dedicated SCSI ID for the
picker device, but accepts both data drive access commands as well as picker
commands through the same SCSI ID. However, the data drive access commands
must be directed to SCSI lu 0 and the picker commands to SCSI lu 1. Therefore, if
you are using the HP SureStore 12000e Exchanger, create a new device file that will
allow you to address the picker through the same SCSI ID as the data drive (ID
number 1), but referring to a different SCSI lu (ID number 1):
mknod /dev/scsi/1.lu1 c 105 0x201110
---|
||
|
|---- Picker SCSI lu (1)
|
----- Picker SCSI ID
|
(the same as data
|
drive SCSI ID = 1)
---------- scsi_pt driver major
9-9
Using Complex Devices and Exchangers
Configuring a SCSI-II Exchanger
3. Using the OmniBack II uma utility verify that the picker identifies itself
correctly as a SCSI-II Exchanger:
# echo “inq” | /usr/omni/bin/uma -ioctl /dev/scsi/0
/dev/scsi/0> inq
----------|
Note:
Use /dev/scsi/1.lu1 for
the HP SureStore 12000e
On HP-UX 10.x systems the uma utility is located in the
/opt/omni/lbin directory.
If the picker does not identify itself correctly, then the selected device file
does not refer to a SCSI-II exchanger picker. Double check that the
picker is referred to by SCSI ID 0. If you are configuring an HP
SureStore 12000e Exchanger, make sure you have created the proper
device file.
4. Using the OmniBack II uma utility to force the exchanger to re-initialize
itself:
# echo "init" | /usr/omni/bin/uma -ioctl /dev/scsi/0
/dev/scsi/0> init
----------|
Note:
Use /dev/scsi/1.lu1
for the HP SureStore 12000e
On HP-UX 10.x systems the uma utility is located in the
/opt/omni/lbin directory.
9-10
Using Complex Devices and Exchangers
Configuring a SCSI-II Exchanger
Configuring the logical device
For our example the following logical devices must be configured:
Logical
Device
Drive
Device File
Control File
DRIVE_1
1
/dev/rmt/1m
/dev/scsi/0
DRIVE_2
2
/dev/rmt/2m
/dev/scsi/0
Table 9-1
Example Device Configuration
1. Once you have determined the correct UNIX device files for the
exchanger drive and picker, configure the SCSI-II Exchanger. Choose
Devices from the OmniBack II Main menu.
2. Select Edit → Create Logical Device → SCSI II Exchanger to open
the Logical Device Editor window.
3. In the Logical Device Name field, enter the logical device name (in
our example, DRIVE_1) and select the associated medium class (in our
example, DAT/DDS).
4. In the Description field, enter descriptive text for this logical device
(for example, “Exchanger Drive 1”).
5. In the Pool Name field, enter the name of the exchanger pool you have
created for your exchanger earlier. Use ... to get a list of available pools.
6. In the Host Name field, enter the name of the computer to which the
exchanger is connected. Use the ... button to choose from a list.
7. Enter the name of an auxiliary logical device (Auxiliary Device) that can
be used if this one is not available at backup run time.
8. In the Exchanger Data Device field, enter the pathname of the Unix
device file corresponding to each exchanger drive. For tape exchangers,
the pathname will be one of the files from the /dev/rmt directory. In our
example, the pathname would be /dev/rmt/1m.
9. In the Drive Index field, enter the ID number of the drive you want
OmniBack II to use for the backup or restore. In our example, "1".
9-11
Using Complex Devices and Exchangers
Configuring a SCSI-II Exchanger
NOTE:
Take care when configuring the logical device that you specify the right device file
for the right drive index. If you do not, you could set it up so that the exchanger loads
media into device 2 and reads from device 1. For a one-drive exchanger, this is
irrelevant (one device file would have the index 1). But for a two- and more device
exchanger this is important.
10. In the SCSI II Control Device field, enter the pathname of the Unix
device file that corresponds to the exchanger’s picker mechanism. In our
example, the pathname of the UNIX device file is /dev/scsi/0.
You do not have to set the Advanced Options button. These options are set to
default values and normally do not need to be adjusted.
This completes the configuration of the first example logical device,
DRIVE_1. To configure the second drive DRIVE_2, follow the same steps.
To Verify the Configuration
From the Media Management window select a range of slots and choose
Edit → Scan...If the device is configured correctly, OmniBack II will be
able to load, read, and unload each medium inside the exchanger repository.
9-12
Using Complex Devices and Exchangers
Configuring a Pre 2.0 SCSI-II Exchanger
Configuring a Pre 2.0 SCSI-II Exchanger
NOTE:
This section describes how to configure SCSI II exchangers using a standard pool, as
supported with previous versions of OmniBack II.
We recommend to configure exchangers using OmniBack II Exchanger Pool. This
allows easier configuration and control of the exchanger device. See “Configuring a
SCSI-II Exchanger” on page 9-7 for instruction.
The generic tape/optical exchanger (also called exchanger, autoloader etc.)
is a complex mechanical device containing one or more data drive
mechanisms, an internal magazine (repository) of media and a mechanical
picker capable of loading, and unloading the media into the drives under
software control.
The generic exchanger has a SCSI ID for each data drive and one for the
picker. For instance, an HP optical jukebox containing two optical drives has
three SCSI IDs — one for the each drive and a third for the picker. One
exception to this is the HP SureStore 12000e Exchanger, which allows both
mechanisms, the data drive and the picker, to be accessed through the same
SCSI ID, but different SCSI Logical Units (lu).
Each repository slot of an exchanger holds one cartridge. OmniBack II
assigns a numerical ID to each slot within the repository starting from one.
You refer to these slots by ID numbers when you configure the logical
device.
For example, an exchanger with 6 repository slots will have the slot IDs 1, 2,
3, 4, 5, and 6. If the exchanger has multiple drives, each drive is referred to
by an ID starting with the number one too. A 4-drive exchanger, for
example, will have the drive IDs 1, 2, 3, and 4. A single drive exchanger is
referred to by the drive ID number “1.”
NOTE:
Note that the above slot and drive IDs are NOT the physical element IDs returned by
the exchanger in response to a SCSI mode sense command (ioctl). The physical
element IDs are determined by the Media Agent based on the above slot and drive
IDs and the mode sense information returned by the exchanger when a backup is run.
9-13
Using Complex Devices and Exchangers
Configuring a Pre 2.0 SCSI-II Exchanger
OmniBack II has provided this numbering scheme so that you do not have to know
the physical element IDs of the exchanger to complete the logical device
configuration.
If you have a multi-drive exchanger and want to use the drives in parallel,
you must configure multiple logical devices, one for each exchanger drive.
You must also decide which set of available repository slots in the exchanger
you want to assign to a single logical device. Take care not to assign the
same repository slot ID to two or more logical devices. This might create
unexpected problems if you use the logical devices concurrently. When
configuring a logical device for an exchanger with one built-in drive only,
you can assign the entire contents of the repository to the logical device.
Example of an Exchanger Configuration
The following example will demonstrate how to configure an exchanger. We
will use a double-drive DAT exchanger with a repository of 50 slots and
connected to an HP9000/S700 system. The steps that follow show you how
to complete the configuration of the hypothetical logical devices, DRIVE_1
and DRIVE_2. We will configure two logical devices for this exchanger, so
we can use both drives in parallel. The first logical device will operate
through drive 1 and the second through drive 2.
We will split the complete repository into two equally sized subsets, the first
containing slot IDs 1 to 25 and the second slot IDs 26 to 50. This will result
in two logical devices which OmniBack II can use in parallel, the operation
of one not interfering with that of the other.
The basic steps involved in configuring the exchanger are:
1. Selecting the device files.
2. Configuring the logical device.
9-14
Using Complex Devices and Exchangers
Configuring a Pre 2.0 SCSI-II Exchanger
Selecting the device files
NOTE:
If you’re configuring your exchanger on an HP9000/S800 system, you will have to
install the SPT driver which is available in the patch PHKL_5405 (for HP-UX 9.04
or later) or PHKL_5406 (for HP-UX 10.x.) Make sure that you follow carefully the
instructions provided with the patch. You must install the correct driver for your
devices, select the correct SCSI address, and create a device file for the device you
will use with OmniBack II.
NOTE:
When this driver is installed and OmniBack II performs backup or restore sessions,
we do not suggest the use of applications running ioscan, such as SAM. This could
cause the system to reboot.
1. Determine the Unix device files that correspond to the exchanger drives.
If you’re configuring a tape exchanger, select the device files from the
/dev/rmt directory.
If you’re using an optical exchanger, select the raw disk device files from
the /dev/rdsk directory.
The double-drive DAT exchanger in our example would use the device
files /dev/rmt/1m and /dev/rmt/2m to address the exchanger drives
(assuming the data drive SCSI IDs are 1 and 2).
2. Determine the UNIX device file that corresponds to the exchanger picker
mechanism.
Since our example uses an HP9000/S700 system, the device file
associated with the exchanger picker would be in the /dev/scsi
directory. We will use the device file /dev/scsi/0. The "0" at the end
of the device file name indicates that SCSI ID for the exchanger picker
device is 0.
9-15
Using Complex Devices and Exchangers
Configuring a Pre 2.0 SCSI-II Exchanger
NOTE:
The HP SureStore 12000e Exchanger does not have a dedicated SCSI ID for the
picker device, but accepts both data drive access commands as well as picker
commands through the same SCSI ID. However, the data drive access commands
must be directed to SCSI lu 0 and the picker commands to SCSI lu 1. Therefore, if
you are using the HP SureStore 12000e Exchanger, create a new device file that will
allow you to address the picker through the same SCSI ID as the data drive (ID
number 1), but referring to a different SCSI lu (ID number 1):
mknod /dev/scsi/1.lu1 c 105 0x201110
---|
||
|
|---- Picker SCSI lu (1)
|
----- Picker SCSI ID
|
(the same as data
|
drive SCSI ID = 1)
---------- scsi_pt driver major
3. Using the OmniBack II uma utility verify that the picker identifies itself
correctly as a SCSI-II Exchanger:
# echo “inq” | /usr/omni/bin/uma -ioctl /dev/scsi/0
/dev/scsi/0> inq
----------|
Note:
Use /dev/scsi/1.lu1 for
the HP SureStore 12000e
On HP-UX 10.x systems the uma utility is located in the
/opt/omni/lbin directory.
If the picker does not identify itself correctly, then the selected device file
does not refer to a SCSI-II exchanger picker. Double check that the
picker is referred to by SCSI ID 0. If you’re configuring an HP SureStore
12000e Exchanger, make sure you have created the proper device file.
4. Using the OmniBack II uma utility to force the exchanger to re-initialize
itself:
# echo "init" | /usr/omni/bin/uma -ioctl /dev/scsi/0
/dev/scsi/0> init
----------|
Note:
Use /dev/scsi/1.lu1
for the HP SureStore 12000e
On HP-UX 10.x systems the uma utility is located in the
/opt/omni/lbin directory.
9-16
Using Complex Devices and Exchangers
Configuring a Pre 2.0 SCSI-II Exchanger
Configuring the logical device
Preparing Media
Before configuring an exchanger to operate with OmniBack II, prepare a
Standard Media Pool for usage with this exchanger. If you don’t want to
use any of the existing or predefined OmniBack II Media Pools, you will
need to configure a new media pool.
In our example, it would be convenient for us to create a separate standard
media pool for each repository slot group in the exchanger. In our example
we would create two pools: pool_DRIVE_1 and pool_DRIVE_2. Pool
pool_DRIVE_1 would keep media residing in slot IDs 1 through 25 (for
logical device DRIVE_1) and pool pool_DRIVE_2 would keep media for
slots 26 through 50 (for logical device DRIVE_2). These pools will be
initially empty. For instructions how to create Media Pools and add media to
the pool see “Media Management” on page 3-13.
Configuring the SCSI II Exchanger
For our example the following logical devices must be configured:
Logical
Device
Drive
Slots
Device File
Control File
DRIVE_1
1
1-25
/dev/rmt/1m
/dev/scsi/0
DRIVE_2
2
26-50
/dev/rmt/2m
/dev/scsi/0
Table 9-2
Example Device Configuration
1. Once you have determined the correct UNIX device files for the
exchanger drive and picker, configure the SCSI-II Exchanger. Choose
Devices from the OmniBack II Admin menu.
2. Select Edit → Create Logical Device → Magazine & Pre 2.0 SCSI
II Exchanger.
3. In the Logical Device field, enter the logical device name (in our
example, DRIVE_1) and select the associated medium class (in our
example, DAT/DDS).
9-17
Using Complex Devices and Exchangers
Configuring a Pre 2.0 SCSI-II Exchanger
4. In the Pool Name field, enter the name of the Media Pool you have
created for this logical device. Do this by clicking on ... to get a list of
available pools. (In our example, select pool pool_DRIVE_1.)
5. In the Host Name field, enter the name of the computer to which the
exchanger is connected. (You can use the ... button to choose from a list.)
6. Enter the name of an auxiliary logical device (Auxiliary Device) that can
be used if this one is not available at backup run time.
NOTE:
You do not have to set the Advanced Options button. These options are set to default
values and normally do not need to be adjusted.
7. In the Description field, enter descriptive text for this logical device
(for example, “Exchanger Drive 1”).
8. In the Exchanger Data Device field, enter the pathname of the Unix
device file corresponding to each exchanger drive. For tape exchangers,
the pathname will be one of the files from the /dev/rmt directory. In our
example, the pathname would be /dev/rmt/1m.
9. In the Drive Index field, enter the ID number of the drive you want
OmniBack II to use for the backup or restore. In our example, "1".
NOTE:
Take care when configuring the logical device that you specify the right device file
for the right drive index. If you don’t, you could set it up so that the exchanger loads
media into device 2 and reads from device 1. For a one-drive exchanger, this is
irrelevant (one device file would have the index 1). But for a two- and more device
exchanger this is important.
10. In the SCSI II Control Device field, enter the pathname of the Unix
device file that corresponds to the exchanger’s picker mechanism. In our
example, the pathname of the UNIX device file is /dev/scsi/0.
11. Enter all repository slot IDs you want to assign to this logical device. You
can specify ranges, for example 1-25 . Click on Add after entering the ID
number range.
This completes the configuration of the first example logical device,
DRIVE_1. You would configure the second logical device DRIVE_2
repeating steps 2 through 11 above, making sure to change all the relevant
fields for DRIVE_2.
9-18
Using Complex Devices and Exchangers
Configuring a Pre 2.0 SCSI-II Exchanger
To Verify the Configuration
Select Edit → Rescan from the OmniBack II Logical Device
Administration GUI. If the device is configured correctly, OmniBack II
should be able to load, read, and unload each medium inside the exchanger
repository.
9-19
Using Complex Devices and Exchangers
Configuring a Non-Standard Exchanger for Use with OmniBack II
Configuring a Non-Standard Exchanger for Use
with OmniBack II
The "External Script Exchanger Policy" module locates the execution of
device specific commands into an external script or program that is called
each time the OmniBack II Media Agent needs to load or unload a medium
from a repository. This interface to the external script (which the user
writes)has been designed so that OmniBack II can operate simply and
efficiently with non-standard exchanger devices. To keep the script interface
as simple as possible but flexible enough to handle a large variety of
exchanger devices, the OmniBack II Media Agent defines only two
operations:
•
medium load
•
medium unload
Both the load and unload operations are invoked through the same external
script which must be capable of parsing our predefined options and
parameters.
Device Configuration
Configuring a logical device involves the following steps:
1. Specifying a logical device name
2. Specifying the media class
3. Selecting a Media Pool
4. Selecting a host
5. Selecting an auxiliary device
6. Entering a logical device description
7. Specifying the exchanger data device
8. Specifying the exchanger control script
9. Specifying the exchanger repository slots
10. Specifying advanced options (optional)
9-20
Using Complex Devices and Exchangers
Configuring a Non-Standard Exchanger for Use with OmniBack II
Configure a logical device using the Logical Device Editor window that
matches the type of logical device you are using.
To access the Logical Device Editor window, follow these steps:
1. At the OmniBack II main menu, select Devices to open the HP
OmniBack II - Logical Device Administration window
appears. This window shows the logical devices that are currently
configured for this cell.
2. Select Edit → Create Logical Device → External Control from the
menu bar. The Logical Device Editor External Control
window appears.
To create a logical device, follow these steps:
1. Enter a logical device name in the Device name field. The device name
can be a maximum of 255 alphanumeric characters containing no spaces.
2. Specify a default Media Pool for this device. Click on the ... button to
select from a list of configured media pools.
3. Specify a host for this device. Click on the ... button to select from a list
of configured hosts.
4. Specify an auxiliary device for this logical device. Click on the ... button
to select from a list of configured auxiliary devices.
5. To specify advanced options for this logical device, click on the
Advanced Options button. This step is optional.
6. Enter a description for this logical device. Use up to 64 alphanumeric
characters including spaces.
7. Enter the complete path and filename of the exchanger data device in the
field labeled Exchanger Data Device. This is the device file through
which the actual data will be read (or written).
8. Enter the Data Driver Index in the field labeled Drive Index. This index
is a number starting from one and must correspond to the Exchanger Data
Device specified in step 7. For single drive exchangers, this index may be
omitted in which case the index defaults to one.
9. Enter the complete path and filename of the Exchanger Control Script in
9-21
Using Complex Devices and Exchangers
Configuring a Non-Standard Exchanger for Use with OmniBack II
the field labeled Exchanger Control Script. This is the name of the
script (program) you have to supply and which OmniBack II will invoke
to load and unload media in the device. (See "Device Usage" in the
following section for the command line parameters passed by OmniBack
II when calling the script.)
10. In the field labeled Exchanger Repository Slots enter the
exchanger repository slot IDs for this logical device, and click the Add
button. Repeat this step for each repository slot contained in this
exchanger. For SCSI-II compliant devices, slots are typically referred to
by numbers starting from 1. However, when you are using the External
Script Exchanger Policy, OmniBack II allows you to use either numbers
or strings for slot addresses. The slot numbers you specify are simply
passed to the Exchanger Control Script.
To Delete a Repository Slot
Select (highlight) the slot number in the list of slot numbers and click the
Delete button.
11. When you have entered all the information for this logical device, click
OK to add this logical device.
Device Usage
At runtime, the OmniBack II Media Agent will call the external script
passing to the script all information necessary to load or unload the media
from the repository. As soon as the script successfully completes the action,
it should return an exit code of zero. If the script terminates unsuccessfully, a
positive integer is returned. During the execution of the script, any messages
read from its stdout will be picked up by the Media Agent and passed to
the controlling Session Manager as error messages at level MINOR. The
script is called by OmniBack II using the popen() mechanism.
NOTE:
Note that using an external script interface is a potential security leak since the
OmniBack II Media Agents are always executed with root permissions.
To load a medium into the designated exchanger drive, OmniBack II invokes
the external script as:
9-22
Using Complex Devices and Exchangers
Configuring a Non-Standard Exchanger for Use with OmniBack II
%<script>
-drive %<data drive index>
-slot
%<one from the list of slot addresses>
[-flip] %<if medium should be flipped during transport>
-load
Similarly, to unload a medium from a drive into its original position,
OmniBack II invokes the script as:
%<script>
-drive %<data drive index>
-slot
%<one from the list of slot addresses>
[-flip] %<if medium should be flipped during transport>
-unload
Note that the OmniBack II Media Agent keeps track of whether the current
medium used is flipped and what its original position in the exchanger
repository slot has been. When the medium is unloaded, it will be flipped
again as it is unloaded.
The Media Agent assumes that the exchanger is on-line and reset to an
operational state before OmniBack II uses it. In a previous section of this
chapter we recommended you always operate the exchanger via the
OmniBack II software. If someone circumvents the software and does
operations on the exchanger manually, the result will be inconsistent
software and exchanger states. If the exchanger is left in this inconsistent
state, the load or unload action will probably fail and abort the Media Agent.
The external script should detect that the exchanger has been used and left in
an inconsistent state and rescan the exchanger so that it is reset to its default
operational state. However, the Media Agent does not issue a special reset or
initialize command for this purpose.
Some exchangers have removable cartridge magazines (containing media)
that can be loaded and unloaded under software control. The OmniBack II
Media Agent assumes the magazine is loaded and does not attempt to preload it at startup or unload it at shutdown. If a particular exchanger offers the
magazine loading function, the external scripts should detect an unloaded
magazine and load it automatically without any Media Agent intervention.
9-23
Using Complex Devices and Exchangers
Configuring a Non-Standard Exchanger for Use with OmniBack II
9-24
10
Troubleshooting
10-1
Troubleshooting
In this chapter
This chapter is divided in sections covering the most common messages and
sections covering other OmniBack II messages. Each message is provided
with description and recommended actions.
10-2
Troubleshooting
Common GUI Errors and Warnings
Common GUI Errors and Warnings
MESSAGE:
Cannot access the server host. (inet is not responding)
The host that is specified in the
/usr/omni/config/cell/cell_server file is not reachable, it is
not turned on and running, or has no OmniBack II software
installed and configured on it.
DESCRIPTION:
The xomni GUI tried to connect to the Cell Server but the connection failed.
ACTION:
Check if the cell server host is up and running. Check if it is accessible on
the network. If not, check the configuration of network services and routing.
Check if the OmniBack II software is properly installed on the host. You can
use the command: telnet hostname 5555 . After 30 seconds the
OmniBack II should respond with the version information to confirm that it
is ready.
MESSAGE:
Cannot export medium medium_label.
Medium is protected.
DESCRIPTION:
You can not export medium with protected data.
ACTIONS:
Recycle medium and then export it.
MESSAGE:
10-3
Troubleshooting
Common GUI Errors and Warnings
Cannot lock record - record is already locked by someone
else.
DESCRIPTION:
Backup is started but someone is modifying logical devices or media
configuration.
You tried to modify logical devices or media configuration but backup is
running.
ACTION:
Wait until backup is completed or until modifications are finished.
MESSAGE:
Cannot remove device. There are still objects backed up
on this device.
ACTIONS:
Before removing the logical device you have to export all media from the
pool.
MESSAGE:
Cannot remove pool.
Pool is used as default pool for device().
DESCRIPTION:
You can not remove pool which is used as the default pool for logical
devices.
ACTION:
Move logical devices to another pool.
10-4
Troubleshooting
Common GUI Errors and Warnings
MESSAGE:
Connect to CRS failed. (OmniBack II daemons are not
running) Use the command omnisv.sh to start the OmniBack
II deamons on the cell server host.
DESCRIPTION:
OmniBack II processes are not running on the Cell Server when the GUI or
command tried to connect.
ACTION:
Use the command omnisv.sh -start to start the processes on the Cell
Server.
Configure OmniBack II automatic startup and shutdown. See the
RELNOTES for instructions.
MESSAGE:
Database overflow. Disabled writing of detail catalog
part.
DESCRIPTION:
The database size limit was exceeded. detail catalog is not written to the
database any more.
ACTION:
•
Purge the database of the obsolete information.
•
Change the database size limit.
MESSAGE:
Device not found.
DESCRIPTION:
When adding/modifying an object to the datalist, you tried to specify logical
device which does not exist.
10-5
Troubleshooting
Common GUI Errors and Warnings
ACTION:
Use help […] button to get list of configured devices.
MESSAGE:
Host hostname unavailable.
Cannot change config files!
DESCRIPTION:
When exporting the host from the cell, it was not reachable. The host was
properly exported only on the Cell Server side.
ACTION:
Check if the host is connected to the network and reachable.
Problem will occur when you will try to import this host to another cell. You
must delete the /usr/omni/config/cell/cell_server file on the
host.
MESSAGE:
Host is not in this cell.
DESCRIPTION:
When adding/modifying an object to the datalist, you tried to get list of
available mountpoints for the host which is not configured in the cell.
ACTION:
Use the ... button to get the list of configured hosts.
Check if the disk agent software is installed on the host.
MESSAGE:
Importing host hostname…
Cannot import host hostname.
10-6
Troubleshooting
Common GUI Errors and Warnings
Command import failed.
DESCRIPTION:
You wanted to import the host to your cell, but the command failed.
ACTION:
Check the /usr/omni/config/cell/cell_server located on this system. This file
contains the information about Cell Server to which this host belongs. If
name of cell server is different, then remove this file and repeat the
procedure.
MESSAGE:
Insufficient disk space on filesystem containing paths
pathname and pathname.
DESCRIPTION:
There is not enough disk space on volumes with /tmp/omni_tmp or
/usr/omni directories.
ACTIONS:
Free up more disk space and start the installation again.
MESSAGE:
MediaAgent for series sun/sparc/sunos is not supported.
DESCRIPTION:
This version of OmniBack II does not support Media Agent for series
sun/sparc/sunos and aix. Check which architectures are supported.
MESSAGE:
Node hostname unknown or not responding.
DESCRIPTION:
Installation on the selected host failed because the host was not reachable.
10-7
Troubleshooting
Common GUI Errors and Warnings
ACTION:
Check if the host you want to install is reachable on the network.
MESSAGE:
OmniBack II medium “Default File_1” in drive.
Medium is still under protection => recycle medium.
Initialization denied.
DESCRIPTION:
If you have an OmniBack II medium and it is still under protection, it cannot
be initialized.
ACTION:
Check that you don’ need data on the medium, recycle it and initialize it
with the Force initialization option.
MESSAGE:
OmniBack II medium “c14d05b9:2fb2f6ff:4a3d:0001” in
drive.
Medium not found in OmniBack II database.
This medium can be initialized only with force option
specified.
Initialization denied.
DESCRIPTION:
If you have an OmniBack II medium and it was not found in the database,
the medium cannot be initialized.
ACTION:
Check that you really do nott need the data on medium any more and
initialize the medium with the Force Initialization option.
MESSAGE:
10-8
Troubleshooting
Common GUI Errors and Warnings
The number of configured clients is greater than the
number of available client (host) licenses. Possibly the
license has expired.
Run command omnicc -query to see the licensing summary
report.
DESCRIPTION:
Not enough licenses.
ACTION:
Add new licenses or change the configuration.
10-9
Troubleshooting
Common Message Screen Errors and Warnings
Common Message Screen Errors and Warnings
MESSAGE:
[Major] From: BMA@hostname “NJIVA_48L” Time: 07/29/95
02:11:53 /dev/dat3
Cannot close device (/ error)
DESCRIPTION:
Something is wrong with the device.
ACTION:
Check if the medium is bad or the device needs cleaning.
MESSAGE:
[Critical] From: BDA-NETIO@hostname “KALEIDOSKOP backup”
Time: Sat Jul 29, 1995 01:33:13
Cannot connect to Media Agent (IPC Cannot Connect) =>
aborting
DESCRIPTION:
Connection to Media Agent failed and cannot be reestablished.
ACTION:
Check the displayed Media Agent messages for some errors.
Try to connect to the Media Agent host using telnet.
10-10
Troubleshooting
Common Message Screen Errors and Warnings
MESSAGE:
[Critical] From: BDA-NETIO@hostname “KALEIDOSKOP backup”
Time: Sat Jul 29, 1995 01:33:13
Cannot connect to Media Agent (IPC Invalid Hostname or
IP Address) => aborting
Unexpected close reading NET message => aborting
DESCRIPTION:
Media Agent host not known on the Cell Server.
ACTION:
Add Media Agent host to the /etc/hosts file or the name server.
MESSAGE:
[Major] From: BMA@hostname “NJIVA_48L” Time: 07/29/95
09:25:43
Cannot load exchanger medium (Function is not available)
DESCRIPTION:
OmniBack II could not load the medium in the exchanger.
ACTION:
Update the exchanger repository. Reset the exchanger. Check if the
configuration of exchanger is valid.
MESSAGE:
Cannot open: (unknown error) => not backed up
DESCRIPTION:
File could not be opened and it was not backed up.
ACTION:
If this error appears constantly, exclude the file from the backup.
10-11
Troubleshooting
Common Message Screen Errors and Warnings
MESSAGE:
[Warning] From: VBDA@hostname Time: 08/14/95 16:34:32
filename
Cannot preserve time attributes: (Not owner)
DESCRIPTION:
OmniBack II could not reset access time attribute of the file.
ACTION:
If this error appears constantly, you may set the Do not preserve time
attributes option, or exclude the file from the backup.
MESSAGE:
[Minor] From: VBDA@hostname “KALEIDOSKOP backup” Time:
07/29/95 03:37:25
filename
Cannot read: (Lockf deadlock detection)
DESCRIPTION:
File is locked by some other process and cannot be backed up with OB-II
ACTION:
If this error appears constantly, exclude the file from the backup.
MESSAGE:
[Minor] From: VBDA@hostname “KALEIDOSKOP backup” Time:
07/29/95 11: 09:50
/users/leon/
Cannot stat: (No such file or directory) => not backed up
DESCRIPTION:
File has been removed while backup was running.
MESSAGE:
10-12
Troubleshooting
Common Message Screen Errors and Warnings
[Minor] From: VBDA@hostname “KALEIDOSKOP backup” Time:
07/28/95 23 :51:51
/nfs/ops
Cannot stat: (Stale NFS file handle) => not backed up
DESCRIPTION:
NFS mount is not valid.
ACTION:
Umount and mount back the specified NFS mounted volume.
MESSAGE:
[Major] From: BMA@hostname “NJIVA_48L” Time: 07/29/95
02:11:53
/dev/dat3
Cannot write to device (/ error)
DESCRIPTION:
Error on the device.
ACTION:
Check if the medium is bad or the device needs cleaning.
MESSAGE:
[Major] From: BSM@hostname “KALEIDOSKOP” Time: 07/29/95
02:17:10
Got unexpected close from Filesystem backup DA on
hostname.
DESCRIPTION:
Disk Agent has lost the connection to the Media Agent. The Media Agent
might be killed or aborted for some reason.
ACTION:
Restart backup.
10-13
Troubleshooting
Common Message Screen Errors and Warnings
MESSAGE:
[Critical] From: VBDA@hostname “KALEIDOSKOP backup”
Time: 07/29/95 03:15:03
Received ABORT request from NET => aborting
DESCRIPTION:
Media Agent was aborted and the Disk Agent must be aborted, too.
ACTION:
Check messages for Media Agent errors.
MESSAGE:
[Major] From: BSM@hostname “ZLATOROG1” Time: 08/12/95
22:12:06
The Filesystem backup DA reached its inactivity timeout
of 7200 seconds.
The agent on host hostname will be shutdown.
DESCRIPTION:
Disk Agent was not active for more than two hours.
ACTION:
Check for mount problems (bdf), big sparse files, or directories with large
number of files…
MESSAGE:
[Critical] From: BMA@hostname “name” Time: 05/11/95
13:08:01
Cannot determine control device type (/dev/scsi/.)
DESCRIPTION:
During the backup session, OmniBack II was no able to open the backup
device.
10-14
Troubleshooting
Common Message Screen Errors and Warnings
ACTION:
Check if the device is switched on, is attached to the hostname, and has the
necessary device file created.
MESSAGE:
[Critical] From: BSM@hostname “name” Time: 05/12/95
15:33:57
The host hostname is needed for the backup session, but
this host is not configured.
Run the command omnicc -query to view the licensing
summary report.
DESCRIPTION:
This message is displayed if you have exported or deinstalled the host, but
the same host is still configured in the datalist.
ACTION:
Modify the datalist.
Reinstall the host or import it.
MESSAGE:
[Critical] From: INET@hostname “name” Time: 05/11/95
12:39:19
User “username” non-existent on this system => aborting
DESCRIPTION:
The backup session on the host was aborted because the user is not
registered on the host.
ACTION:
OmniBack II requires that the user has the same username on all hosts you
back up.
10-15
Troubleshooting
Common Message Screen Errors and Warnings
MESSAGE:
[Critical] From: VBDA@hostname “name” Time: 05/11/95
12:40:07 tree
Not valid mount point => aborting
DESCRIPTION:
Wrong mountpoint for the hostname was configured, the backup was
aborted.
ACTION:
Modify the object and use the […] button to get list of valid mountpoints for
the host.
Check available filesystems using the UNIX df command.
MESSAGE:
[Major] From: BSM@hostname “name” Time: 05/11/95
13:08:05
Could not connect to inet for starting of disk agent on
host “hostname”.
DESCRIPTION:
During the backup session the agent on the host hostname could not be
started and the host was not backed up.
ACTION:
Check if the host is up and running. Check if it is accessible on the network.
Check if the OmniBack II software is properly installed on the host. You can
use the telnet hostname 5555. After 30 seconds the OmniBack II should
respond with the version information to confirm that it is running.
MESSAGE:
[Minor] From: BMA@hostname “name” Time: 05/11/95
13:51:59
Cannot load medium, source slot () appears to be empty
10-16
Troubleshooting
Common Message Screen Errors and Warnings
DESCRIPTION:
OmniBack II tries to use new media first. The slot is empty.
ACTION:
Insert medium in the slot and continue.
MESSAGE:
[Minor] From: BSM@hostname “name” Time: 05/11/95
10:58:36
There are not enough media in the pool to complete the
backup.
DESCRIPTION:
The calculation at the end of the backup preview showed that there is not
enough free media in the media pool.
ACTION:
Add new media to the pool.
Check if the media has the correct size defined.
If some backups are no longer needed, you may want to recycle media to
create more space for new backups.
MESSAGE:
[Minor] From: VBDA@hostname “name” Time: 05/11/95
10:54:40 tree
Cannot traverse: (Permission denied) => not backed up
DESCRIPTION:
You do not have UNIX permissions to back up the specific tree. The backup
of that tree was skipped.
ACTION:
Exclude the tree form the backup as you are not authorized to backup these
files.
10-17
Troubleshooting
Common Message Screen Errors and Warnings
MESSAGE:
[Warning] From: VBDA@hostname “name” Time: 05/11/95
10:57:25
tree
Object does not reside on current filesystem => not
backed up
DESCRIPTION:
The tree you have specified is not on that filesytem.
ACTION:
Check on which filesystem the tree resides and modify the datalist
accordingly.
MESSAGE:
rexec: Lost connection
Problem in obtaining information from system hostname.
Possibly wrong password.
DESCRIPTION:
Installation to the new system failed- system function rexec between Cell
Server and client host cannot be executed.
ACTION:
Resolve why the rexec function cannot be executed. See man pages for
rexec command.
10-18
Troubleshooting
Less Common Warning and Error Messages
Less Common Warning and Error Messages
MESSAGE:
[ERROR] Bad session ID.
CAUSE:
Command received an ID that is not valid or doesn’ represent running
session.
ACTION:
Run omnistat command to see all session IDs or use the GUI.
MESSAGE:
[ERROR] Can not open database/file.
CAUSE:
OmniBack II was unable to open the database. The database files are either
not available or the Unix file permissions to access the database files are
incorrect. This error can occur if the omnidbcheck or omnidbinit
command is used by user other than root. Only root should use these two
commands.
ACTION:
Check the database files (omni.dbd, *.dat, *.ext, *.key) in the
‘/usr/omni/config/db’’ directory on the Cell Server. The permissions of these
files should be 644 and user root:other. If there is only the omni.dbd
file, database has not been initialized yet. In this case, use the omnidbinit
command to initialize the database.
MESSAGE:
[ERROR] Can’ access the file.
10-19
Troubleshooting
Less Common Warning and Error Messages
CAUSE:
configuration file cannot be accessed. This message means that the cell
server daemon can’ read or write from certain configuration file.
ACTION:
Use the /bin/ll - command to check the appropriate files below the
directory /usr/omni/config.
MESSAGE:
[ERROR] Can’ access the file.
CAUSE:
The IPC communication ended (unexpectedly).
ACTION:
Restart operation.
MESSAGE:
[ERROR] Can’ get the lock for file. The file is being
accessed by some other OmniBack II process or some
program was killed not unlocking the file. In that case
lock will be re-taken in 30 minutes.
CAUSE:
The Session Manager was unable to lock the database for update. Either the
database is being updated by another Session Manager or program was
killed (locking the database). This Session Manager will try to again at minute intervals to lock the database.
ACTION:
Stop the Cell Server (using the omnisv.sh -stop command) and restart
the Cell Server (using the omnisv.sh -start command). This will
release all database locks.
10-20
Troubleshooting
Less Common Warning and Error Messages
MESSAGE:
[MAJOR] FullPathName Cannot append to medium
(UnixErrorString)
CAUSE:
The Backup Media Agent (BMA) process was unable to append to existing
data on the medium in use.
ACTION:
No action is necessary. OmniBack II will attempt to use another medium as
this medium has no available space. Simply respond to any mount prompts.
MESSAGE:
[WARNING] FullPathName Cannot apply lock:
(UnixErrorString)
CAUSE:
The specified object could not be locked before backup. The error message
indicates the reason as reported by the operating system.
ACTION:
•
•
Eliminate the problem described by the error message.
Unset the "Lock Files During Backup" option in the configuration of the
backup object.
MESSAGE:
[MINOR] FullPathName Cannot close: (UnixErrorString)
CAUSE:
If the specified object can’ be closed after it is backed up, it’ backup status is
unsuccessful. The error message indicates the reason as reported by the
operating system.
ACTION:
•
Eliminate the problem described by the error message and restart the
10-21
Troubleshooting
Less Common Warning and Error Messages
•
backup session.
Exclude the specified object from the backup fileset by using the exclude option.
MESSAGE:
[ERROR] Cannot connect to host hostname.
CAUSE:
OmniBack II could not connect to the specified host.
ACTION:
Use the pin command to check the connection to this host.
MESSAGE:
[ERROR] Cannot connect to required session session-id.
CAUSE:
You specified the wrong session ID when trying to abort the session.
ACTION:
Use the omnistat command to get the correct session ID, then try the abort
again.
MESSAGE:
[ERROR] Cannot connect to session with session ID
session-id.
CAUSE:
You tried to monitor session which is not backup or restore session. You can
only monitor backup and restore sessions.
ACTION:
Restart the monitor, specifying backup or restore session.
10-22
Troubleshooting
Less Common Warning and Error Messages
MESSAGE:
[MAJOR] FullPathName Cannot determine disk
characteristics (UnixErrorString) => aborting
CAUSE:
The optical pre-erase operation failed because OmniBack II was unable to
read the disk.
ACTION:
•
•
Verify that the disk drive model is supported by OmniBack II.
Verify that the disk is not damaged. Replace any damaged disks.
MESSAGE:
[MAJOR] FullPathName Cannot erase disk surface
(UnixErrorString) => aborting
CAUSE:
The pre-erase operation failed due to system error. The error message
indicates the reason as reported by the operating system.
ACTION:
Eliminate the problem described by the error message and restart the
session. If the pre-erase fails again, you should replace the disk.
MESSAGE:
[ERROR] Cannot execl omnib for datalist ’datalist_name’.
CAUSE:
The omnitrig command can’ execute the command omnib.
ACTION:
10-23
Troubleshooting
Less Common Warning and Error Messages
Check to make sure that the /usr/omni/bin/omnib command exists, and
make sure that the Unix file permissions are correctly set. If the omnib
command exists and the permissions are correct, the file may be corrupt. In
this case, reinstall the Cell Console on the Cell Server.
MESSAGE:
[WARNING] FileName Cannot force lock: Unix Error Code
CAUSE:
The restore operation could not lock the specified object before starting the
restore. The error message indicates the reason as reported by the operating
system.
ACTION:
Restart the operation.
MESSAGE:
[WARNING] FullPathName Cannot force lock:
(UnixErrorString)
CAUSE:
The specified object could not be locked before backup. The error message
indicates the reason as reported by the operating system.
ACTION:
•
•
Eliminate the problem described by the error message.
Toggle off the "Lock Files During Backup" option of the backup.
MESSAGE:
[ERROR] Cannot install trigger in cron. Cannot open
crontab for writing.
CAUSE:
10-24
Troubleshooting
Less Common Warning and Error Messages
You either are not configured for cron or do not have permission to access
the cron file.
ACTION:
Check, and if necessary correct, the cron file configuration and the
permission of its configuration files.
MESSAGE:
[MAJOR] Cannot load exchanger medium (UnixErrorString)
CAUSE:
The exchanger failed to complete SCSI move command. The error message
indicates the reason as reported by the operating system.
ACTION:
Verify that the autochanger device is online and ready. If necessary, perform
hardware reset of the autochanger or turn the power off and on. Then restart
the failed session.
MESSAGE:
[MINOR] Cannot load medium, source slot
(RepositorySlotID) appears to be empty
CAUSE:
The exchanger status indicates that the source slot in the exchanger
repository does not contain movable medium.
ACTION:
If the exchanger does contain movable medium, the internal exchanger
status is corrupted and should be reset. To fix the problem, terminate the
OmniBack II session, reset the exchanger and rescan the exchanger using
the OmniBack II rescan feature.
NOTE:
10-25
Troubleshooting
Less Common Warning and Error Messages
You can reset the exchanger by either pressing the reset button on the front
panel or turning the power off and on.
MESSAGE:
[ERROR] Cannot lock database - currently unavailable.
CAUSE:
You tried to execute process that requires exclusive access to the database,
but the database is locked by another process.
ACTION:
Check to make sure that no other OmniBack II processes are running on the
Cell Server (especially session managers (BSM/RSM/MSM/DBSM) and the
omnidbcheck command.) Stop the Cell Server (using the omnisv.sh stop command) and then restart the Cell Server (using the omnisv.sh start command).
MESSAGE:
[MAJOR] FullPathName Cannot lock exchanger control
device (UnixErrorString)
CAUSE:
The Media Agent (xMA) could not reserve the exchanger control device for
its exclusive use. Exclusive locking is required for multi-drive exchangers to
prevent mechanical damage. This problem can occur if you’re running
several Media Agents that concurrently use the same autochanger. The error
message indicates the reason as reported by the operating system.
ACTION:
Check the device and restart the session.
MESSAGE:
[MAJOR] Cannot lock medium mediumID in database.
Probably in use by someone else. Skipping erase.
10-26
Troubleshooting
Less Common Warning and Error Messages
CAUSE:
You tried to erase medium which is locked by another process. Possible
causes are: another user is modifying medium data (label and location) or
someone killed the backup/restore Session Manager and the medium stayed
locked.
ACTION:
Wait until the medium is free and try again.
MESSAGE:
[MAJOR] Cannot lock medium mediumID in database.
Probably in use by someone else. Skipping reinitialization.
CAUSE:
You tried to reinitialize medium which is locked by another process.
Possible causes are: another user is modifying medium data (label and
location) or someone killed the backup/restore Session Manager and the
medium stayed locked.
ACTION:
Wait until the medium is free and try again.
MESSAGE:
[MINOR] FileName Cannot mknod : Unix Error Code
CAUSE:
You must have the access right for restore as root to restore special
files.
ACTION:
Either add the restore as root access right to this user class or transfer
this user to user class which has the restore as root access right.
10-27
Troubleshooting
Less Common Warning and Error Messages
MESSAGE:
[ERROR] Cannot open database in exclusive mode. Some
session managers may be still running.
CAUSE:
You tried to run process which opens the database in exclusive mode.
However, OmniBack II was unable to open the database in exclusive mode
because some Session Managers were still executing.
ACTION:
Wait for the session managers to complete, then restart the operation.
MESSAGE:
[ERROR] Cannot open database. Somebody is running
omnidbcheck command or OmniBack II database backup in
progress!
CAUSE:
You tried to run process which opens the database. However, OmniBack II
was unable to open the database because either user is running
omnidbcheck or an OmniBack II database backup was in progress.
ACTION:
Wait for omnidbcheck or database backup to complete, then restart the
operation.
MESSAGE:
[MAJOR] FullPathName Cannot open device
(UnixErrorString)
CAUSE:
The Media Agent (xMA) detected an error trying to open the device. The
error message indicates the reason as reported by the operating system. "No
such device or address" message from media device indicates that there is no
medium in the device.
10-28
Troubleshooting
Less Common Warning and Error Messages
ACTION:
Verify that medium is loaded and ready. Respond to any pending mount
prompts.
MESSAGE:
[MAJOR] Cannot open mount table: (MountTable)
CAUSE:
OmniBack II is unable to access the register of mounted filesystems.
Probably invalid access rights to the register.
ACTION:
Verify the cause of this problem described in the error message and notify
your system administrator.
MESSAGE:
MAJOR: Cannot open mount table: (UnixErrorString)
CAUSE:
OmniBack II is unable to access the register of mounted filesystems on the
system running the disk agent (VBDA). The error message indicates the
reason as reported by the operating system.
ACTION:
Verify that the mount table (file /etc/mnttab on HP-UX systems) exists
on the system running the disk agent (VBDA) and can be accessed.
MESSAGE:
[WARNING] FullPathName Cannot open option file:
(UnixErrorString)
CAUSE:
10-29
Troubleshooting
Less Common Warning and Error Messages
The Disk Agent (xBDA) received -read option or an input redirection
operator (-trees %<filename) and the input file could not be accessed.
ACTION:
Verify that the input file exists on the appropriate host and can be read by the
Disk Agent (xBDA).
MESSAGE:
[WARNING] FullPathName Cannot open option file:
(UnixErrorString)
CAUSE: The Media Agent (xMA) received -read option or an input
redirection operator (-dev %<file) and the input file could not be accessed.
ACTION:
Verify that the input file exists on the appropriate host and can be read by the
Media Agent (xMA).
MESSAGE:
[MINOR] FullPathName Cannot open: (UnixErrorString) =>
not backed up
CAUSE:
The specified object could not be opened and will not be backed up. The
error message indicates the reason as reported by the operating system.
ACTION:
•
•
Eliminate the problem described by the error message and restart the
backup session.
Exclude the specified object from the backup fileset using the -exclude
option.
MESSAGE:
10-30
Troubleshooting
Less Common Warning and Error Messages
[WARNING] FullPathName Cannot preserve time attributes:
(UnixErrorString)
CAUSE:
The disk agent (VBDA) is unable to preserve the time attributes of the
specified object. The error message indicates the reason as reported by the
operating system.
ACTION:
Toggle off the "Restore Time Attributes" option of the restore.
MESSAGE:
[MINOR] FullPathName Cannot read ACLs: (UnixErrorString)
CAUSE:
The extended ACL attributes belonging to this object cannot be read and
will not be backed up. The error message indicates the reason as reported by
the operating system.
ACTION:
Eliminate the cause of the problem described by the error message and
restart the backup session to complete successful backup.
MESSAGE:
[MAJOR] FullPathName Cannot read from device
(UnixErrorString)
CAUSE:
The Media Agent (xMA) detected an error while reading from the device.
The error message indicates the reason as reported by the operating system.
ACTION:
The session fails. You should check the media in the device and restart the
session.
10-31
Troubleshooting
Less Common Warning and Error Messages
MESSAGE:
[MINOR] FullPathName Cannot read link: (UnixErrorString)
=> not backed up
CAUSE:
The specified symbolic link cannot be resolved and will not be backed up.
The error message indicates the reason as reported by the operating system.
ACTION:
Eliminate the cause of the problem described by the error message and
restart the backup session to complete successful backup.
MESSAGE:
[MINOR] FullPathName Cannot read: (UnixErrorString)
CAUSE:
The specified object could not be read and will not be backed up. The error
message indicates the reason as reported by the operating system.
ACTION:
•
•
Eliminate the problem described by the error message and restart the
backup session.
Exclude the specified object from the backup fileset using the -exclude
option.
MESSAGE:
[ERROR] Cannot recycle medium, because there are private
objects on it.
CAUSE:
You tried to recycle medium containing backup sessions having private
objects.
ACTION:
10-32
Troubleshooting
Less Common Warning and Error Messages
To reuse this medium, you must first reinitialize the medium using the
force initialization option.
MESSAGE:
[MAJOR] FullPathName Cannot seek to requested position
(UnixErrorString)
CAUSE:
The Media Agent (xMA) detected an error while trying to move the
read/write head to certain position. The error message indicates the reason as
reported by the operating system.
ACTION:
possible cause for this error is corrupted media format. Verify the media
format using the OmniBack medium verification utility.
The session fails. You should check the media in the device and restart the
session.
MESSAGE:
[MINOR] FullPathName Cannot stat: (UnixErrorString) =>
not backed up
CAUSE:
The specified object could not be examined and will not be backed up. The
error message indicates the reason as reported by the operating system.
ACTION:
•
•
Eliminate the problem described by the error message and restart the
backup session.
Exclude the specified object from the backup fileset by using the exclude option.
MESSAGE:
10-33
Troubleshooting
Less Common Warning and Error Messages
[MINOR] FullPathName Cannot traverse: (UnixErrorString)
=> not backed up
CAUSE:
The specified directory could not be traversed (probably due to insufficient
access rights) and will not be backed up. The error message indicates the
reason as reported by the operating system.
ACTION:
•
•
Eliminate the problem described by the error message and restart the
backup session.
Exclude the specified directory from the backup fileset using the exclude option.
MESSAGE:
[MAJOR] Cannot unload exchanger medium (UnixErrorString)
CAUSE:
The exchanger failed to complete SCSI move command. The error message
indicates the reason as reported by the operating system.
ACTION: Verify that the autochanger device is online and ready. If
necessary, perform hardware reset of the autochanger or turn the power off
and on. Then restart the failed session.
MESSAGE:
[MINOR] Cannot unload medium, source drive (DriveSlotID)
appears to be empty
CAUSE:
The exchanger status indicates that the source drive is empty. An empty
drive cannot be unloaded.
ACTION:
10-34
Troubleshooting
Less Common Warning and Error Messages
If there is medium in the source drive, then the internal exchanger status is
corrupted and needs to be reset. To fix the problem, terminate the OmniBack
II session, reset the exchanger and rescan the exchanger using the
OmniBack II rescan feature.
NOTE:
You can reset the exchanger by either pressing the reset button on the front
panel or turning the power off and on.
MESSAGE:
[MINOR] Cannot unload medium, target slot
(RepositorySlotID) appears to be occupied
CAUSE:
The exchanger status indicates that the target slot in the exchanger
repository already contains medium. The unload operation fails and the
Media Agent (xMA) terminates.
ACTION:
If the source drive is empty, then the internal exchanger status is corrupted
and needs to be reset. To fix the problem, terminate the OmniBack II session,
reset the exchanger and rescan the exchanger using the OmniBack II rescan
feature.
NOTE:
You can reset the exchanger by either pressing the reset button on the front
panel or turning the power off and on.
MESSAGE:
[WARNING] FullPathName Cannot unlock exchanger control
device (UnixErrorString)
CAUSE:
The Media Agent (xMA) could not release the exchanger control device.
The error message indicates the reason as reported by the operating system.
10-35
Troubleshooting
Less Common Warning and Error Messages
ACTION:
No action is necessary. The exchanger control device will be released
automatically as soon as the Media Agent (xMA) process terminates.
MESSAGE:
[WARNING] FileName Cannot unlock: Unix Error Code
CAUSE:
The restore operation could not unlock the specified object after the restore,
therefore the restore failed. The error message indicates the reason as
reported by the operating system.
ACTION:
Restart the restore.
MESSAGE:
[WARNING] FullPathName Cannot unlock: UnixErrorString
CAUSE:
The specified object could not be unlocked after backup and it’ backup
status will not be successful. The error message indicates the reason as
reported by the operating system.
ACTION:
No action is necessary. The specified object will be unlocked automatically
when the disk agent (VBDA) process terminates.
MESSAGE:
[MAJOR] Cannot update repository of non-exchanger
device.
CAUSE:
10-36
Troubleshooting
Less Common Warning and Error Messages
You tried to rescan device which is not an autochanger. You can only rescan
an autochanger device.
ACTION:
Select an autochanger device and restart the rescan.
MESSAGE:
[ERROR] Cannot use un-allocated medium by strict pool
policy.
CAUSE:
You did not mount the medium requested by the mount prompt. If you are
using pool which has strict Media Allocation Policy, you must mount the
exact medium specified by the mount request.
ACTION:
Mount the medium specified by the mount prompt.
MESSAGE:
[MAJOR] FileSystemMountPoint Cannot verify filesystem
mount point:
CAUSE:
OmniBack II is unable to locate the specified mount point in the register of
mounted filesystems.
ACTION:
•
•
Verify that the specified mount point exists on the system running the disk
agent (VBDA) and can be accessed.
Ensure that the specified filesystem is mounted at backup time
MESSAGE:
[MAJOR] FullPathName Cannot write to device
UnixErrorString
10-37
Troubleshooting
Less Common Warning and Error Messages
CAUSE:
The Media Agent (xMA) detected an error while writing to the device. The
error message indicates the reason as reported by the operating system.
ACTION:
The session fails. You should check the media in the device and restart the
session.
MESSAGE:
[WARNING] Connection to the peer_type on host peer_host
is inactive for over than the timeout period and will
therefore be closed. Agent identification is
peer_version.
CAUSE:
Trouble communicating with one of the agents. Agent does not update its
status periodically.
ACTION:
Restart the operation.
MESSAGE:
[MAJOR] Could not connect to Disk Agent
agent_host:mountpoint named label.
CAUSE:
It is not possible to start Disk Agent on specified host. There might be
problem with your network setup or there is no OmniBack II inet process
properly configured for the target host.
ACTION:
Check your network setup and check that the configuration of the files
/etc/services and /etc/inetd.conf on the target host is correct for
the inet entry.
10-38
Troubleshooting
Less Common Warning and Error Messages
MESSAGE:
[MAJOR] Could not connect to Media Agent
agent_host:mountpoint named label.
CAUSE:
It is not possible to start medium agent on specified host. There might be
problem with your network setup or there is no OmniBack II inet process
properly configured for the target host.
ACTION:
Check your network setup and check that the configuration of the files
/etc/services and /etc/inetd.conf on the target host is correct for
the inet entry.
MESSAGE:
[MAJOR] Could not connect to OmniBack II inet process on
host media_agent_host.
CAUSE:
Media management Session Manager cannot connect to OmniBack II inet
process on reported host.
Possible reasons are:
•
•
•
OmniBack II is not installed on that host
The /etc/inetd is not running
The inetd configuration is incorrect
ACTION:
Make sure that the OmniBack II entries are contained in the
/etc/inetd.conf and/etc/services files. Then trigger the
/etc/inetd daemon to reread the configuration with /etc/inetd -.
MESSAGE:
[MAJOR] Could not start Media Agent. Aborting…
10-39
Troubleshooting
Less Common Warning and Error Messages
CAUSE:
Could not start the Media Agent for Init or List Media. The Media Agent did
not respond to the Session Manager. Check the message from OmniBack II
inet process.
ACTION:
Make sure that the OmniBack II software is installed on host where this
device is located.
MESSAGE:
[ERROR] Database is probably corrupted. Please run
omnidbcheck!
CAUSE:
OmniBack II database is probably corrupt.
ACTION:
Terminate all running sessions and GUI’ so that no process is accessing the
database. Next, login to Cell Server as root and run the omnidbcheck fix command. If you get error message "Database is corrupted!", the
database is unfixable and you must recover the database from backup.
MESSAGE:
[ERROR] Database key files have been corrupted. Please
run omnidbcheck -fix!
CAUSE:
The OmniBack II database index file has been corrupted.
ACTION:
Use the omnidbcheck -fix command to fix the index file.
MESSAGE:
10-40
Troubleshooting
Less Common Warning and Error Messages
[ERROR] Database network communication error.
CAUSE:
The OmniBack II database has problems communicating with the lockmgr.
ACTION:
Check to make sure that the lock manager (lockmgr) is running. Stop the
Cell Server (using the omnisv.sh -stop command) and restart the Cell
Server (using the omnisv.sh -start command).
MESSAGE:
[ERROR] Database record limit per file reached.
CAUSE:
The limit for OmniBack II database records has been reached.
ACTION:
Stop all Session Managers and terminate all GUI’, then use the
omnidbcheck -purge command to delete obsolete data from the database.
Then restart the operation.
MESSAGE:
[MAJOR] Default pool pool_name of specified device is
invalid.
CAUSE:
The pool you selected for this device is invalid because it uses an
incompatible media class.
ACTION:
Configure different pool for this device.
MESSAGE:
10-41
Troubleshooting
Less Common Warning and Error Messages
[MAJOR] Device device_name and pool pool_name media
classes do not match. Aborting…
CAUSE:
The media class of this device is different than the media class of the
specified pool.
ACTION:
Select pool which has the same media class as the device.
MESSAGE:
[MAJOR] Error accessing the database, in line code_line,
file code_file. Database layer reports:
"error_description".
CAUSE:
Problems with the OmniBack II database.
ACTION:
Use the omnidbcheck to check the OmniBack II database.
MESSAGE:
[ERROR] File containing Cell Server host not found.
CAUSE:
The configuration file /usr/omni/config/cell/cell_server which
holds the name of the cell server host is missing.
ACTION:
Create the configuration file /usr/omni/config/cell/cell_server
and add the name of the local Cell Server.
MESSAGE:
[ERROR] File containing cell server host not found.
10-42
Troubleshooting
Less Common Warning and Error Messages
CAUSE:
The file /usr/omni/config/cell/cell_server is missing. This file
contains information about the Cell Console and the Cell Server.
ACTION:
Create this file and add the name of the local Cell Server.
MESSAGE:
[WARNING] Filesystem (FileSystem) is READ ONLY !
CAUSE:
FileSystem on which you want to restore data is mounted as read-only.
ACTION:
To fix the problem, follow these steps:
1. Umount the filesystem
2. Remount the filesystem as read/write filesystem
3. Restart the restore
MESSAGE:
[WARNING] FilesystemMountPoint Filesystem is mounted
read-only => locking disabled
CAUSE:
The filesystem being backed up is mounted as read-only or is read-only by
nature (CDROM) and does not support mandatory locking. The disk agent
(VBDA) will disable the locking feature automatically to prevent wrong
error messages.
ACTION:
Toggle off the "Lock Files During Backup" option of the backup.
MESSAGE:
10-43
Troubleshooting
Less Common Warning and Error Messages
[MAJOR] Filesystem too deep: (FilesystemRecursionDepth)
levels
CAUSE:
The Filesystem Backup Disk Agent has reached the maximum depth of its
recursive filesystem traversal. This indicates that your directory structure is
too deep.
ACTION:
•
•
Try to reconfigure the filesystem backup object to contain several trees
starting at lower level in the directory hierarchy.
Decrease the depth of the directory structure to less than 50 levels.
MESSAGE:
[ERROR] Handshaking error. Versions of client and Cell
Server don’ match.
CAUSE:
Internal compatibility check reported that the running GUI or command is
not compatible with the running cell server daemon.
ACTION:
Check the OmniBack II release on the hosts and update the software.
MESSAGE:
[ERROR] Instance of OmniBack II Maintenance control
script already running or terminated abnormally;
semaphore file inst.lock will be ignored if no other
instance running…
CAUSE:
The lock file inst.lock already existed when the installation was started.
ACTION:
10-44
Troubleshooting
Less Common Warning and Error Messages
If you are sure that only you are running the installation script, you can
proceed with the installation.
MESSAGE:
[ERROR] Internal error: Invalid Ct function argument
specified.
CAUSE:
OmniBack II database layer internal error.
ACTION:
Check the database using the omnidbcheck command. If the database is
corrupt, restore the backed up OmniBack II database.
MESSAGE:
[MAJOR] Interprocess communication error on Session
Manager connection. The connection was to peer_type on
host peer_host. Agent identification is “peer_version”.
Ipc library reports "detailed_error_description".
CAUSE:
Unexpected error for interprocess communication on connection to an agent
or monitor. The connection is closed. If the connection was to monitor you
can ignore this message.
ACTION:
Restart the operation.
MESSAGE:
[MAJOR] Invalid cartridge slotID specified.
CAUSE:
You specified slot which is not configured for this device.
10-45
Troubleshooting
Less Common Warning and Error Messages
ACTION:
Check the device configuration and specify the correct slot.
MESSAGE:
[MAJOR] Invalid device device_name specified.
CAUSE:
This cell does not contain the device you specified.
ACTION:
Check the spelling of the device name and the configuration of the device.
MESSAGE:
[MAJOR] Invalid format of the message from peer_type on
host peer_host. Agent identification is “peer_version”.
CAUSE:
One of the agents received known message with an invalid format.
ACTION:
Restart the session.
MESSAGE:
[WARNING] Invalid option “OptionString” => ignored
CAUSE:
The disk agent (xBDA) received an invalid option from the Session
Manager. It is possible that the user specified an option argument starting
with dash character (-).
ACTION:
•
•
Check that option arguments do not have leading dash.
Verify that the Disk Agent (xBDA) and backup Session Manager (BSM)
10-46
Troubleshooting
Less Common Warning and Error Messages
versions are compatible.
MESSAGE:
[MAJOR] Invalid version of peer peer_id, version
peer_version.
CAUSE:
Reported peer (Media Agent or command/GUI) is not compatible with this
version of the Session Manager.
ACTION:
Check the software version of your cell and update the peer to the same
release as the Session Manager.
MESSAGE:
[ERROR] Lock daemon error - probably not running.
CAUSE:
The OmniBack II database lock manager (lockmgr) is not running.
ACTION:
Check to make sure that the lock manager (lockmgr) is running. Stop the
Cell Server (using the omnisv.sh -stop command) and restart the Cell
Server (using the omnisv.sh -start command).
MESSAGE:
[ERROR] Maximum number of session managers allowed are
currently running.
CAUSE:
Too many concurrent sessions running.
ACTION:
10-47
Troubleshooting
Less Common Warning and Error Messages
Use the monitor GUI or the omnistat command to check the number of
currently executing sessions. Abort any unnecessary sessions. Stop the Cell
Server (using the omnisv.sh -stop command) and restart the Cell Server
(using the omnisv.sh -start command). This will terminate all active
GUI’ and command-line interfaces.
MESSAGE:
[MAJOR] Media Agent failed! Aborting…
CAUSE:
The connection to Media Agent was lost. Either the Media Agent was
aborted or wrong shared libraries are in /usr/omni/lib.
ACTION:
Check the installation and version of the shared libraries
/usr/omni/lib/libpol.sl and /usr/omni/lib/libdev.sl.
MESSAGE:
[MAJOR] Media allocation failed for the pool with strict
allocation policy. Media agent aborted.
CAUSE:
The automatic media allocation policy failed. Check the following possible
causes:
•
•
No media remaining in your current media pool
Some media remained locked due to abnormal backup termination
ACTION:
If the pool is empty, add media to the pool or use different pool.
MESSAGE:
[WARNING] Medium already in database. You should export
medium first.
10-48
Troubleshooting
Less Common Warning and Error Messages
CAUSE:
The medium you tried to import already belongs to this cell.
ACTION:
To put the medium in different cell, use the Move Medium feature.
Otherwise, first export medium then import the medium.
MESSAGE:
[ERROR] Medium currently in use by someone else.
CAUSE:
You tried to use medium which is locked by either an executing process or
process which terminated unsuccessfully.
ACTION:
If other session managers are running, you can’ use this medium until the
other Session Managers finish (or are terminated).
MESSAGE:
[WARNING] Medium is still under protection => recycle
medium.
CAUSE:
You tried to write to medium which contains protected sessions.
ACTION:
Use the OmniBack II Recycle feature to remove the protection for the
medium, then restart the operation.
MESSAGE:
[MAJOR] FullPathName Medium not in OmniBack II format =>
aborting
CAUSE:
10-49
Troubleshooting
Less Common Warning and Error Messages
The medium examined is not in OmniBack II format and therefore cannot be
verified by OmniBack II.
ACTION:
Mount valid OmniBack II medium and restart the verification.
MESSAGE:
[WARNING] Medium of type medium_type discovered. Use of
the medium denied.
CAUSE:
OmniBack II checks the format of the medium before writing to the
medium. OmniBack II recognizes media of the type such as tar, cpio, and
HP-UX filesystem, but will not write to these media formats.
ACTION:
To use media of these formats, you must first initialize the medium by using
the -force option. An alternative is to use UNIX commands to overwrite
the first block on the medium.
MESSAGE:
[ERROR] Monitoring is not supported for this session
type.
CAUSE:
You tried to monitor session which is not backup or restore session. You can
only monitor backup and restore sessions.
ACTION:
Restart the monitor, specifying backup or restore session.
MESSAGE:
[ERROR] Neither /etc/hosts nor Name Service can
resolve/verify system hostname.
10-50
Troubleshooting
Less Common Warning and Error Messages
CAUSE:
Installation failed because the host name was not in the /etc/hosts file.
ACTION:
Enter the hostname again, making sure that you spell the name correctly.
MESSAGE:
[ERROR] No details for this session type.
CAUSE
You tried to view status information for session which is not backup or
restore session. You can only view status information for backup and restore
sessions.
ACTION:
Retry the operation, specifying backup or restore session.
MESSAGE:
[ERROR] No session manager with that session ID is
currently running.
CAUSE:
There is no session with that session ID currently running.
ACTION:
Use omnistat to display all currently running sessions.
MESSAGE:
[MAJOR] No side of cartridge for double-sided device
specified.
CAUSE:
10-51
Troubleshooting
Less Common Warning and Error Messages
You selected device which uses double-sided media but you did not specify
which side of the medium to use.
ACTION:
Specify the side of the medium to use ( or ).
MESSAGE:
[MAJOR] No slot ID for exchanger device specified.
CAUSE:
You have selected an autochanger device and must therefore specify slot ID.
ACTION:
Specify slot ID for the operation (Init/erase/list/import/verify operations).
MESSAGE:
[ERROR] Not enough disk space for database.
CAUSE:
OmniBack II didn’ have enough disk space to use the database.
ACTION:
Delete or move files from the disk where the OmniBack II database is
located (/usr/omni/config/db) then restart the operation.
MESSAGE:
[ERROR] Not enough memory for using database.
CAUSE:
OmniBack II was unable to open the database because there was not enough
memory.
ACTION:
10-52
Troubleshooting
Less Common Warning and Error Messages
Either close or kill some processes on Cell Server machine to make more
memory available, then restart the operation.
MESSAGE:
[WARNING] FullPathName Not in hard link catalog =>
backed up regularly
CAUSE:
hard link reference was created after the Disk Agent (xBDA) completed its
first treewalk building its internal hard link catalog. The object will be
backed up as regular file and cannot be restored as hard link reference, but as
separate file only.
ACTION:
If you are concerned about the consistence of your backup data, restart the
backup session.
MESSAGE:
[ERROR] Nothing to browse or database was closed during
browse due to inactivity.
CAUSE:
There are two possible causes:
•
•
You tried to use the Browse feature to search the OmniBack II database
for files. OmniBack II was unable to display this information because the
OmniBack II lacks detailed information about the backup session you
selected. Possible causes are: the backup session was performed without
the Log files to database option or the backup session failed.
The browser window remained inactive for more than 30 minutes.
ACTION:
Try the browser again. If the cause was inactivity, the browser will work. If
the cause was lack of detailed information in the OmniBack II database, you
will need to using another method (for example, read the media) to locate
the file().
10-53
Troubleshooting
Less Common Warning and Error Messages
MESSAGE:
[WARNING] FullPathName Object does not reside on current
filesystem => not backed up
CAUSE:
The specified object (file/directory) does not belong to the physical
filesystem being backed up. Probably it is an NFS mount point.
ACTION:
No action is necessary. To suppress this error message, exclude the specified
object from the backup fileset by using the -exclude option.
MESSAGE:
[WARNING] FullPathName Object is mounted filesystem
CAUSE:
You have specified rawdisk section pathname which refers to disk device
containing filesystem currently mounted.
ACTION:
No actions is necessary. If you are concerned about the consistence of your
backup data, you should use the pre/post-exec feature to unmount the
filesystem before the backup and re-mount it after the backup has
completed.
MESSAGE:
[MAJOR] deviceFileName Object is mounted filesystem
CAUSE:
The rawdisk has mounted filesystem.
ACTION:
Unmount the filesystem and start the restore again.
10-54
Troubleshooting
Less Common Warning and Error Messages
MESSAGE:
[MAJOR] FullPathName Object is not character device =>
not backed up
CAUSE:
You have specified rawdisk section pathname which does not refer to
character device file.
ACTION:
Verify the configuration of the backup object and the rawdisk section files on
the host running the disk agent (RBDA).
MESSAGE:
[MAJOR] Object version not found in the database.
Database layer reports: "error_description".
CAUSE:
Cannot find this information in the database.
ACTION:
Read the media by using the List->From Media feature of the OmniBack
II restore.
MESSAGE:
[ERROR] Parameter string is too long.
CAUSE:
Action response is too long (exceeds the internal limits)
ACTION:
This message means that the output of low level query is bigger than the
limit. Split the command in more commands in order to use less resources.
MESSAGE:
10-55
Troubleshooting
Less Common Warning and Error Messages
[ERROR] Problems in transfer of package packet-name to
system hostname via service-name service. Check service
manually for user username before continuing.
CAUSE:
FTP failed while transferring OmniBack II software packet to the selected
host. The FTP failed because user used the su command instead of su -.
ACTION:
Run su - before you start the installation.
MESSAGE:
[ERROR] Protected medium and cannot append.
CAUSE:
You tried to append to medium which is unappendable either because the
medium belongs to pool which is not appendable or because this medium is
not first medium of the backup.
ACTION:
Make sure that the pool is appendable. Then load the first medium of that
backup set.
MESSAGE:
[WARNING] Script mount_prompt_script for the off line
mount request notification dispatching not found.
System error reported is "error_description".
CAUSE:
Unable to locate this mount prompt script. Note: Script must be owned by
root. The script can only be executed if no other user has write permission
for the script.
ACTION:
Check the script and, if necessary, change the write permissions.
10-56
Troubleshooting
Less Common Warning and Error Messages
MESSAGE:
[WARNING] Session already found in database. Skipping
medium.
CAUSE:
There are two possible causes:
•
•
You tried to import medium containing backup session which is already
in the database for this cell.
You tried to import medium from different cell which has session with the
same sessionID as session already in this database.
ACTION:
Export all media for the session and then retry the import.
MESSAGE:
[MAJOR] Session manager received close from the
peer_type on host peer_host The agent did not respond in
time. Connection will be closed. Agent identification is
“peer_version”.
CAUSE:
The connection to the Disk Agent, Media Agent or Monitor was closed
unexpectedly. This is probably due to abnormal agent termination or agent
host shutdown.
ACTION:
Restart the operation.
MESSAGE:
[ERROR] Session with session ID session-id is not
running.
CAUSE:
10-57
Troubleshooting
Less Common Warning and Error Messages
You specified an incorrect session ID, therefore OmniBack II couldn’
connect to the session or display the status of the session.
ACTION:
Use the omnistat command to get the correct session ID, then retry the
operation.
MESSAGE:
[MAJOR] Specified device for the object not found.
CAUSE:
OmniBack II cannot find the specified device.
ACTION:
Check the spelling of the device name and the device configuration.
MESSAGE:
[MAJOR] Specified pool pool_name is invalid.
CAUSE:
You specified nonexistent pool.
ACTION:
Select one of the configured pools for this device.
MESSAGE:
[ERROR] System belongs to different packet-name
installation with Cell Server cell-server-name instead
of cell-server-name.
CAUSE:
10-58
Troubleshooting
Less Common Warning and Error Messages
The distribution of OmniBack II modules to this host failed because the host
already belongs to another OmniBack II cell. The installation script checks
the omni_info file located in the /usr/omni/bin/install directory on the host
where you wanted to install the OmniBack II modules.
ACTION:
If you want to import the selected host into your cell you should first export
it from the cell it belongs to now. Then you will be able to include it into
your cell.
MESSAGE:
[ERROR] There are still objects backuped on this device.
CAUSE:
You tried to delete logical device, but the database still contains objects
which were backed up using this logical device.
ACTION:
The backup history for this object may be obsolete. See PURGING for
instructions how to remove obsolete data.
MESSAGE:
[WARNING] This medium can be initialized only with force
option specified.
CAUSE:
You tried to initialize medium which contained protected data.
ACTION:
Restart the initialization, using the -force option.
MESSAGE:
[ERROR] Unknown event event received.
10-59
Troubleshooting
Less Common Warning and Error Messages
CAUSE:
This problem is caused by an internal error.
ACTION:
Restart the operation.
MESSAGE:
[MAJOR] Unknown event on Session Manager from peer_type
on host peer_host. Agent identification is
“peer_version”.
CAUSE:
Unknown event received from one of the agents.
ACTION:
Restart the session.
MESSAGE:
[WARNING] FullPathName Unknown object type => not backed
up
CAUSE:
The stat() system call identifies filesystem object not recognized and/or
supported by OmniBack II.
ACTION:
No actions are necessary. To suppress this error message, exclude the
specified object from the backup fileset by using the -exclude option.
MESSAGE:
[WARNING] Unknown type of compression :
UniqueDataCompressionType
CAUSE:
10-60
Troubleshooting
Less Common Warning and Error Messages
There is an error in the shared library libdc.sl. There is difference
between the compression used for the backup and the compression used for
the restore. Since OmniBack II was unable to decompress the file, it restored
the file in compressed form.
ACTION:
Make sure that the shared library contains the same compression used for the
backup. Then restart the restore.
MESSAGE:
[WARNING] Unknown type of compression :
uniqueDataCompressionType
CAUSE:
There is an error in shared library libde.sl. There is difference between
the compression used for the backup and the compression used for the
restore. Since OmniBack II was unable to decompress the file, it restored the
file in compressed form.
ACTION:
Make sure that the shared library contains the same compression used for the
backup. Then restart the restore.
MESSAGE:
[WARNING] Unknown type of encryption !
CAUSE:
There is an error in shared library libde.sl. There is difference between
the encryption used for the backup and the encryption used for the restore.
Since OmniBack II was unable to de-encrypt the file, it restored the file in
encrypted form.
ACTION:
Make sure that the shared library contains the same encryption used for the
backup. Then restart the restore.
10-61
Troubleshooting
Less Common Warning and Error Messages
MESSAGE:
[WARNING] Unrecognized option "OptionString" => ignored
CAUSE:
The Media Agent (xMA) received an invalid option from the Session
Manager. It is possible that the user specified an option argument starting
with hyphen -.
ACTION:
•
•
Check that option arguments do not have leading hyphen -.
Verify that the Media Agent (xMA) and Session Manager (xSM) versions
are compatible.
MESSAGE:
[WARNING] Wrong medium loaded.
CAUSE:
You loaded non-OmniBack II medium into the drive.
ACTION:
Either load an OmniBack II medium, or initialize the medium and then use
it.
MESSAGE:
[ERROR] You cannot abort session of this type.
CAUSE:
You tried to abort session which is not backup or restore session. You can
only abort backup and restore sessions.
ACTION:
Wait for the session to end.
MESSAGE:
10-62
Troubleshooting
Less Common Warning and Error Messages
[ERROR] You don’ have permission to abort session.
CAUSE:
OmniBack II security has reported that you have no permission to abort the
session. If user belongs to an OmniBack II class that has the Abort ACL
disabled he will not be able to abort sessions belonging to other users.
ACTION:
Set the Abort ACL in the appropriate class.
MESSAGE:
[ERROR] You don’ have permission to monitor session.
CAUSE:
OmniBack II security has reported that you have no permission to monitor
this session. If user belongs to an OmniBack II class that has the Monitor
ACL disabled he will not be able to monitor sessions belonging to other
users.
ACTION:
Set the Monitor ACL in the appropriate class.
MESSAGE:
[ERROR] You don’ have permission to reply to mount
request.
CAUSE:
OmniBack II security has reported that you have no permission to respond to
the specified mount request. This occurs when you try to respond to mount
request from session that was not initiated by yourself and you belong to
group that has the ACL Mount Prompt disabled.
ACTION:
Set the Mount Prompt ACL in the appropriate class.
10-63
Troubleshooting
Less Common Warning and Error Messages
MESSAGE:
[MAJOR] [BlockSegment:BlockOffset] Invalid block header
=> not in OmniBack II format
CAUSE:
The block at the specified position is not in OmniBack II format. This can
occur if an existing OmniBack II medium has been partly overwritten by
another application.
ACTION:
In order to use this medium for OmniBack II, you must first re-initialize the
medium. To discard the medium, first export it from the OmniBack II pool.
MESSAGE:
[MAJOR] [BlockSegment:BlockOffset] Invalid record header
=> block corrupted
CAUSE:
The block of the OmniBack II media format at the specified position
contains an invalid record header which indicates that the block might be
corrupted.
ACTION:
No recovery actions are possible. Use different medium.
MESSAGE:
[MAJOR] [BlockSegment:BlockOffset] Stored/Computed CRC
mismatch => block corrupted
CAUSE:
The CRC checksum which was computed at backup time does not match the
CRC computed during the verify session. The block at the specified position
is probably corrupted.
ACTION:
10-64
Troubleshooting
Less Common Warning and Error Messages
It is rare that random bit errors can produce this message. Therefore, you
should export this medium from the OmniBack II database and discard the
medium.
MESSAGE:
[COMMAND ERROR] Cell Server host file not found.
CAUSE:
The file /usr/omni/config/cell/cell_server is missing.
ACTION:
Create the cell server file and enter in the name of the local cell server.
MESSAGE:
[COMMAND ERROR] No valid hostname found.
CAUSE:
There was no valid hostname found in the cell server file
/usr/omni/config/cell/cell_server.
ACTION:
Enter the name of your local cell server into the cell server file.
MESSAGE:
[COMMAND ERROR] You don’ have permission to restore on
other hosts.
CAUSE:
You tried to perform restore to another host but you have no permission to
do this.
ACTION:
10-65
Troubleshooting
Less Common Warning and Error Messages
Either add the restore to other hosts access right to this user class or
transfer this user to user class which has the restore to other hosts
access right.
MESSAGE:
[COMMAND ERROR] You don’ have permission to start
restore.
CAUSE:
You tried to perform restore but you have no permission to perform restore.
ACTION:
Either add the start restore access right to this user class or transfer this
user to user class which has the start restore access right.
MESSAGE:
[COMMAND ERROR] You have no permission to start backup.
CAUSE:
You tried to start backup but you have no permission to do this.
ACTION:
Either add the start backup access right to this user class or transfer this
user to user class which has the start backup access right.
MESSAGE:
[COMMAND ERROR] You have no permission to start datalist
backup.
CAUSE:
You tried to start datalist backup but you have no permission to do this.
ACTION:
10-66
Troubleshooting
Less Common Warning and Error Messages
Either add the start datalist backup access right to this user class or
transfer this user to user class which has the start datalist backup
access right.
MESSAGE:
[COMMAND] Cannot start media management session manager
on host hostname.
CAUSE:
You tried to perform media management operation, but you lack the Media
configuration access right.
ACTION:
Either add the Media configuration access right to this user class or
transfer this user to user class which has the Media configuration
access right.
MESSAGE:
[COMMAND] Could not fork process
command
CAUSE:
The command you specified could not execute, for one of several reasons:
•
•
•
the command was not able to connect to the selected Cell Server
the command is not contained in the /usr/omni/bin directory
there are too many processes are already running
ACTION:
Check the following items to correct the problem:
1. Use the omnisv.sh -status command to make sure that the Cell
Server is running.
2. Check the /usr/omni/bin directory to make sure that the command is
in the directory.
3. Use the ps -ef command to make sure that the process limit of the
10-67
Troubleshooting
Less Common Warning and Error Messages
system has not been reached.
MESSAGE:
[COMMAND] You have no permission to get logical device
information.
CAUSE: You tried to view logical device information, but you lack the
Device configuration access right.
ACTION:
Either add the Device configuration access right to this user class or
transfer this user to user class which has the Device configuration
access right.
MESSAGE:
[COMMAND] You have no permission to get media management
information.
CAUSE:
You tried to view media management information, but you lack the Media
configurations access right.
ACTION:
Either add the Media configuration access right to this user class or
transfer this user to user class which has the Media configuration
access right.
MESSAGE:
[COMMAND] You have no permission to initialize medium.
CAUSE:
You tried to initialize medium, but you lack the Media configuration
access right.
ACTION:
10-68
Troubleshooting
Less Common Warning and Error Messages
Either add the Media configuration access right to this user class or
transfer this user to user class which has the Media configuration
access right.
MESSAGE:
[CRITICAL] No physical device() specified => aborting
CAUSE:
No physical devices have been specified for this logical device. Each logical
device must have at least one physical device.
ACTION:
Verify the configuration of the logical device, and if necessary, modify the
logical device.
MESSAGE:
[CRITICAL] Incorrect type of backup - backup must be
done with: typeOfBackup
CAUSE:
You tried to restore file using the wrong menu selection.
ACTION:
To restore file from rawdisk backup, use the Restore Single File
from Rawdisk menu selection.
MESSAGE:
[CRITICAL] FilesystemMountPoint Not valid mount point =>
aborting
CAUSE:
The filesystem mount point specified by the user is not real mount point on
the system running the disk agent (VBDA).
10-69
Troubleshooting
Less Common Warning and Error Messages
ACTION:
Verify the configuration of the backup object and the table of mounted
filesystems on the host running the disk agent (VBDA).
MESSAGE:
[CRITICAL] FullPathName Cannot open exchanger control
device (UnixErrorString)
CAUSE:
The Media Agent (xMA) could not open the specified exchanger control
device. The error message indicates the reason as reported by the operating
system.
ACTION:
Verify that the control device file exists on the appropriate host and can be
accessed by the Media Agent (xMA) process.
MESSAGE:
[CRITICAL] FullPathName Cannot preerase (invalid
physical device type) => aborting
CAUSE:
You can only pre-erase magneto-optical media. The device you selected is
not the correct device type.
ACTION:
Attempt pre-erase operation on magneto-optical media only. Note that not
all HP optical drives are capable of pre-erasing an optical platter.
MESSAGE:
[CRITICAL] FullPathName Cannot preerase (unknown medium
capacity) => aborting
CAUSE:
10-70
Troubleshooting
Less Common Warning and Error Messages
The optical pre-erase operation failed because the capacity of the medium
could not be determined.
ACTION:
Do not use the pre-erase feature on this optical disk. OmniBack II generally
requires that the capacity of disk is known in order to use and handle it
properly.
NOTE:
If only some of the optical disks within your autochanger refuse to be preerased, this may be an indication that their surfaces are damaged. You should
replace the defective disks.
MESSAGE:
[CRITICAL] FullPathName FATAL error opening device
(UnixErrorString)
CAUSE:
The Media Agent (xMA) detected fatal error while trying to open the device.
The error message indicates the reason as reported by the operating system.
ACTION:
Verify the configuration of the logical device and if necessary, modify the
logical device.
MESSAGE:
[CRITICAL] FullPathName Invalid exchanger control device
=> aborting
CAUSE:
The exchanger control device accessed device which is not SCSI-II
compliant exchanger.
ACTION:
10-71
Troubleshooting
Less Common Warning and Error Messages
Verify the configuration of the logical device. The exchanger control device
file probably does not refer to an autochanger picker device.
MESSAGE:
[CRITICAL] FullPathName Invalid physical device type =>
aborting
CAUSE:
The physical device type does not match the device type of the logical
device as configured.
ACTION:
Verify the device type of the physical device and the logical device
configuration.
MESSAGE:
[CRITICAL] FullPathName Object is not disk device => not
backed up
CAUSE:
You have specified rawdisk section pathname which does not refer to disk
device.
ACTION:
Verify the configuration of the backup object and the relevant device files on
the host running the disk agent (RBDA).
MESSAGE:
[CRITICAL] ShellCommandLine Cannot popen() script:
(UnixErrorString)
CAUSE:
10-72
Troubleshooting
Less Common Warning and Error Messages
The Media Agent (xMA) could not invoke subshell script using the
popen() system call. The error message indicates the reason as reported by
the operating system.
ACTION:
Verify that the shell command line has the correct syntax and does not
contain specify any inaccessible or non-executable commands.
Verify that the system running the Media Agent (xMA) agent has sufficient
resources to the necessary processes.
MESSAGE:
[CRITICAL] ShellCommandLine Error reading script output
(UnixErrorString)
CAUSE:
The Disk Agent (xBDA) process detected an error while reading from the
pipe through which it is connected with its pre-/post-exec process. The error
is probably caused by an ungraceful termination of the script. The error
message indicates the reason as reported by the operating system.
ACTION:
•
•
Verify that the shell command line is syntactically correct and does not
contain inaccessible or non-executable command invocations.
Verify that the executed programs behave as expected and do not produce
unexpected results. Test the shell command line manually for possible
bugs.
MESSAGE:
[CRITICAL] ShellCommandLine Error reading script output
(UnixErrorString)
CAUSE:
10-73
Troubleshooting
Less Common Warning and Error Messages
The Media Agent (xMA) process detected an error while reading from the
pipe through which it is connected with its subshell process. The cause is
probably an ungraceful termination of the subshell. The error message
indicates the reason as reported by the operating system.
ACTION:
Verify that the shell command line has the correct syntax and does not
contain any inaccessible or non-executable commands.
Verify that the executed programs work as expected and do not produce
unexpected results. Test the shell command line manually for possible bugs.
MESSAGE:
[CRITICAL] UserName Not Super-User => cannot restore raw
CAUSE:
You must have the restore as root access right to restore rawdisk or
files from rawdisk.
ACTION:
Toggle on the access right for restore as root
MESSAGE:
[CRITICAL] deviceFileName Object is not disk => not
restored
CAUSE:
The specified filename is not character device file.
ACTION:
Check the restore object configuration.
MESSAGE:
10-74
Troubleshooting
Less Common Warning and Error Messages
[CRITICAL] Cannot determine control device type
(FullPathName)
CAUSE:
The device accessed by the specified exchanger control device does not
respond correctly to the SCSI inquiry command. The error message
indicates the reason as reported by the operating system.
ACTION:
Verify the configuration of the logical device. The exchanger control device
file does probably not refer to valid SCSI device.
MESSAGE:
[CRITICAL] Cannot determine repository configuration
(FullPathName)
CAUSE:
The device accessed by the specified exchanger control device does not
respond correctly to the SCSI element address command. The Media Agent
(xMA) cannot continue since it is unable to determine the addresses of the
physical repository elements. The error message indicates the reason as
reported by the operating system.
ACTION:
Verify the configuration of the logical device. The exchanger control device
file does probably not refer to an autochanger picker device. If the file is
correct, verify that the autochanger device is online and ready.
MESSAGE:
[CRITICAL] Cannot execute ’FullPathName’
(UnixErrorString) => aborting
CAUSE:
The inet agent is unable to execute the requested agent. The error message
indicates the reason as reported by the operating system.
10-75
Troubleshooting
Less Common Warning and Error Messages
ACTION:
•
•
Verify that HP OmniBack II is installed correctly on the system.
Verify that the specified agent binary can be accessed and executed on the
system.
MESSAGE:
[CRITICAL] Cannot fork NETIO process (UnixErrorString)
=> aborting
CAUSE:
The Disk Agent (xBDA) main process is not able to fork its NETIO
subprocess. This is possibly due to insufficient system resources on the
system running the Disk Agent (xBDA) agent. The error message indicates
the reason as reported by the operating system. The most probable cause is
"insufficient memory" or "process table full".
ACTION:
Verify that the system running the Disk Agent (xBDA) agent has sufficient
resources to fork and execute another process.
MESSAGE:
[CRITICAL] Cannot fork NETIO process (UnixErrorString)
=> aborting
CAUSE:
The Media Agent (xMA) main process is not able to fork its NETIO
subprocess. possible cause is insufficient system resources on the system
running the Media Agent (xMA) agent. The error message indicates the
reason as reported by the operating system. The most probable cause is
"insufficient memory" or "process table full".
ACTION:
Verify that the system running the Media Agent (xMA) agent has sufficient
resources to fork and execute another process.
10-76
Troubleshooting
Less Common Warning and Error Messages
MESSAGE:
[CRITICAL] Cannot initialize pipe connection
(UnixErrorString) => aborting
CAUSE:
The Disk Agent (xBDA) parent process is not able to open pipe connection
and fork its NETIO process. This is possibly due to insufficient system
resources on the system running the Disk Agent (xBDA) agent. The error
message indicates the reason as reported by the operating system.
ACTION:
Verify that the system running the Disk Agent (xBDA) agent has sufficient
resources to open pipe and establish an IPC connection.
MESSAGE:
[CRITICAL] Cannot initialize pipe connection
(UnixErrorString) => aborting
CAUSE:
The Media Agent (xMA) parent process is not able to open pipe connection
and fork its NETIO process. possible cause is insufficient system resources
on the system running the Media Agent (xMA) agent. The error message
indicates the reason as reported by the operating system.
ACTION:
Verify that the system running the Media Agent (xMA) agent has sufficient
resources to open pipe and establish an IPC connection.
MESSAGE:
[CRITICAL] Cannot load medium, target drive
(DriveSlotID) appears to be busy
CAUSE:
10-77
Troubleshooting
Less Common Warning and Error Messages
The exchanger status indicates that the target drive is not empty. The
medium cannot be loaded into an occupied drive and therefore the Media
Agent (xMA) terminates.
ACTION:
Verify that the drive of the logical device is empty before starting OmniBack
II. If the drive is empty, then the internal exchanger status is corrupted and
needs to be reset. To fix the problem, terminate the OmniBack II session,
reset the exchanger and rescan the exchanger using the OmniBack II rescan
feature.
NOTE:
You can reset the exchanger by either pressing the reset button on the front
panel or turning the power off and on.
MESSAGE:
[CRITICAL] Database layer reports [line: code_line,
file: code_file]: ”error_message”
CAUSE:
The backup session tries to update the database with information, but the
information is not accepted. The probable cause is an invalid transaction. In
rare cases, this error could indicate corrupted database.
ACTION:
Restart the backup. If the error reoccurs, use the omnidbcheck command to
check the database.
MESSAGE:
[CRITICAL] Final BMA acknowledgment missing => backup
INCOMPLETE
CAUSE:
10-78
Troubleshooting
Less Common Warning and Error Messages
The Disk Agent (xBDA) was disconnected from the Media Agent before the
last pending data block was acknowledged. This indicates that the backup
record stream is not complete and the backup cannot be regarded as
successful.
ACTION:
The backup will not be completed successfully. You should restart the
backup session.
MESSAGE:
[CRITICAL] IPC failure reading NETIO message
(UnixErrorString) => aborting
CAUSE:
The Media Agent (xMA) main process detected an IPC error reading an
incoming NETIO message. The error message indicates the reason as
reported by the operating system.
ACTION:
The session will fail. You should restart the session.
MESSAGE:
[CRITICAL] IPC failure reading SM message
(UnixErrorString) => aborting
CAUSE:
The Media Agent (xMA) main process detected an IPC error while reading
an incoming Session Manager (SM) message. The error message indicates
the reason as reported by the operating system.
ACTION:
The session fails. You should restart the session.
MESSAGE:
10-79
Troubleshooting
Less Common Warning and Error Messages
[CRITICAL] Incorrect type of backup - backup must be
done with: TypeOfBackup
CAUSE:
You selected "Restore File from Rawdisk" but the backup object you
selected was not rawdisk.
ACTION:
To restore from rawdisk, you must select "rawdisk" backup object. Either
select rawdisk object or select "Restore File from Filesystem."
MESSAGE:
[CRITICAL] Invalid device type specified
(LogicalDeviceTypeID) => aborting
CAUSE:
The device type configured for this logical device is invalid.
ACTION:
Verify the configuration of the logical device, and if necessary, modify the
logical device.
MESSAGE:
[CRITICAL] Invalid exec code (AgentExecCode) => aborting
CAUSE:
Internal Error: The Session Manager tried to execute an invalid or
nonexistent agent.
ACTION:
Verify that the inet and Session Manager versions are compatible.
MESSAGE:
10-80
Troubleshooting
Less Common Warning and Error Messages
[CRITICAL] Invalid storage element requested => aborting
Please verify logical device configuration
CAUSE:
The set of repository identifiers contains invalid items which do not match
the actual repository elements located inside the exchanger. The error was
detected while attempting to access this invalid element.
ACTION:
Verify the layout of the exchanger repository and the configuration of the
logical device.
MESSAGE:
[CRITICAL] MA/NETIO protocol error => aborting
CAUSE:
The Media Agent (xMA) main process received an unexpected message
type on its NETIO IPC channel.
ACTION: The session fails. You should restart the session.
MESSAGE:
[CRITICAL] No datalist in Session Manager command line
options or no such datalist exists.
CAUSE:
Either an invalid datalist or no datalist was specified.
ACTION:
Specify valid datalist and restart the backup.
MESSAGE:
[CRITICAL] No exchanger control device specified =>
aborting
10-81
Troubleshooting
Less Common Warning and Error Messages
CAUSE:
No SCSI-II exchanger control device has been specified for this logical
device. SCSI-II exchanger must have control device to handle medium
move commands.
ACTION:
Verify the configuration of the logical device and if necessary, modify the
logical device.
MESSAGE:
[CRITICAL] Protocol error, cannot parse options =>
aborting
CAUSE:
Internal Error: The startup handshake protocol between the OmniBack inet
agent and the invoking Session Manager is unsuccessful.
ACTION:
Verify that the inet and Session Manager versions are compatible.
MESSAGE:
[CRITICAL] Received ABORT request from NETIO => aborting
CAUSE:
The NETIO process has requested an immediate abort. This is most
probably caused by an unexpected disconnect event on the connection to the
Media Agent, indicating an ungraceful termination of the Media Agent.
ACTION:
The backup will not be completed successfully. You should restart the
backup session.
MESSAGE:
10-82
Troubleshooting
Less Common Warning and Error Messages
[CRITICAL] SM/MA protocol error => aborting
CAUSE:
The Media Agent (xMA) main process received an unexpected message
type on its Session Manager (SM) IPC channel.
ACTION:
Verify that the Media Agent (xMA) and Session Manager (xSM) versions
are compatible.
The session fails. You should restart the session.
MESSAGE:
[CRITICAL] Security violation: Cannot switch to uid= in
user mode
CAUSE:
The user root can only be member of the user class Admin. This is to
prevent network security problems. The error was detected by the inet
agent enforcing the security.
ACTION:
Verify that user ’root’ is only member of the Admin user class.
MESSAGE:
[CRITICAL] The peer_type on host peer_host did not
identify itself on startup. Connection will be closed.
CAUSE:
Agent must first identify itself to Session Manager in order to check
compatibility. The problem is probably due to the inet process on the target
host being unable to execute the proper agent.
ACTION:
Check the availability of OmniBack II executables on the target host, their
permissions and the availability of shared libraries.
10-83
Troubleshooting
Less Common Warning and Error Messages
MESSAGE:
[CRITICAL] Unexpected close reading NETIO message
(UnixErrorString) => aborting
CAUSE:
The Media Agent (xMA) main process detected loss of connection while
reading an incoming NETIO message. The error message indicates the
reason as reported by the operating system.
ACTION:
The session fails. You should restart the session.
MESSAGE:
[CRITICAL] Unexpected close reading SM message
(UnixErrorString) => aborting
CAUSE:
The Media Agent (xMA) main process detected connection loss while
reading an incoming Session Manager (SM) message. The error message
indicates the reason as reported by the operating system.
ACTION:
The session fails. You should restart the session.
MESSAGE:
[CRITICAL] Unexpected error parsing NETIO message
(UnixErrorString) => aborting
CAUSE:
The Media Agent (xMA) main process detected parser error while reading
an incoming NETIO message. This indicates that the received message was
incorrectly formatted.
ACTION:
The session fails. You should restart the session.
10-84
Troubleshooting
Less Common Warning and Error Messages
MESSAGE:
[CRITICAL] Unexpected error parsing SM message
(UnixErrorString) => aborting
CAUSE:
The Media Agent (xMA) main process detected parse error while reading an
incoming Session Manager (SM) message. This indicates that the received
message was incorrectly formatted.
ACTION:
The session fails. You should restart the session.
MESSAGE:
[CRITICAL] Unknown type of encryption !
CAUSE:
There is an error in shared library libde.sl. There is difference between
the encryption used for the backup and the encryption used for the restore.
Since OmniBack II was unable to de-encrypt the file, it restored the file in
encrypted form.
ACTION:
Make sure that the shared library contains the same encryption used for the
backup. Then restart the restore.
MESSAGE:
[CRITICAL] Unrecognized option "option" !
CAUSE:
The Session Manager sent an invalid option to the Disk Agent (xRDA). This
reason could be that the user specified tree (or similar) with leading dash "-".
ACTION:
10-85
Troubleshooting
Less Common Warning and Error Messages
Check that there are no leading dashes"-". Also, check that the Disk Agent
(xRDA) version corresponds to the Session Manager.
MESSAGE:
[CRITICAL] User "UnixUserName" non-existent on this
system => aborting
CAUSE:
HP OmniBack II requires user to have valid Unix "user:group" account in
order to start an agent on system. The error was detected by the inet agent
enforcing the security.
ACTION:
Either create Unix account for the user on the system, or run the agents in
root mode.
MESSAGE:
[CRITICAL] Agent peer_id, version peer_version, is not
compatible with current version of Session Manager.
Connection will be closed.
CAUSE:
The agent is not compatible with the current version of the Session Manager.
ACTION:
Install the new version of the agent on the hosts. To update all the installed
agents on all the nodes in cell, use the Update All feature of the
OmniBack II Install feature.
MESSAGE:
[CRITICAL] FullPathName Cannot compute disk capacity
(UnixErrorString) => not backed up
10-86
Troubleshooting
Less Common Warning and Error Messages
CAUSE:
OmniBack II is unable to determine the capacity of the specified rawdisk
section. The section cannot be backed up.
ACTION:
The disk is not supported by HP OmniBack II.
MESSAGE:
[CRITICAL] Error in the database consistency discovered.
The database will not be backed up. Checking procedure
reports: "detailed_error_description". For more
information please run omnidbcheck.
CAUSE:
OmniBack II internal database is not in consistent state, so it will not be
backed up.
ACTION:
Run the omnidbcheck -fix command to check the consistency of the
database and rebuild the files.
MESSAGE:
[CRITICAL] Group "UnixGroupName" non-existent on this
system => aborting
CAUSE:
HP OmniBack II does not permit user "user:group" to start an agent on
system where his Unix group does not exist. The error was detected by the
inet agent enforcing the security.
ACTION:
Either create Unix group for the user on the system or change the
configuration of this user to valid Unix group.
10-87
Troubleshooting
Less Common Warning and Error Messages
MESSAGE:
[CRITICAL] User "UnixUserName" is not member of group
"UnixGroupName" => aborting
CAUSE:
HP OmniBack II does not permit Unix user ("user:group") to start an agent
on system where he is not member of the specified Unix group. The error
was detected by the inet agent enforcing the security.
ACTION:
Either create Unix group for the user on the system or change the
configuration of this user to valid Unix group.
MESSAGE:
[CRITICAL] ShellCommandLine Cannot popen() script:
(UnixErrorString)
CAUSE:
The Disk Agent (xBDA) cannot invoke pre-/post-exec script using the
popen() system call. The error message indicates the reason as reported by
the operating system.
ACTION:
•
•
Verify that the shell command line is syntactically correct and does not
contain inaccessible or non-executable command invocations.
Verify that the computer running the Disk Agent (xBDA) has sufficient
resources to spawn another process.
MESSAGE:
[ERROR GUI] Table allocation error. (No more free
entries!).
CAUSE:
Table allocation error.
10-88
Troubleshooting
Less Common Warning and Error Messages
ACTION:
Internal program static tables are used due to several error recovery
procedures. Restart the GUI.
MESSAGE:
[ERROR in OpC integration] Cannot get value of the
environment variable DISPLAY.
CAUSE:
You tried to get the status of specified cell, but OmniBack II was unable to
display this information.
ACTION:
Possibles causes are:
•
•
Make sure that the DISPLAY environment variable is set
Make sure that the OpC management station allows the Cell Server node
to display an hpterm (you must allow access to the appropriate client). To
allow access to the client, use the xhost + clientname command.
MESSAGE:
[ERROR in OpC integration] No valid host running cell
server daemon located.
CAUSE:
You selected node which is not defined as Cell Server node.
ACTION:
Define the node as Cell Server node using the procedures described in the
chapter "OmniBack II and OperationsCenter Integration".
MESSAGE:
10-89
Troubleshooting
Less Common Warning and Error Messages
[NORMAL] No reply from peer_type on host peer_host.
Agent identification is peer_version. The connection to
the agent will be closed.
CAUSE:
The Session manager was waiting for reply from Disk Agent or Media
Agent. The reply did not arrive before the timeout.
ACTION:
Restart the operation.
MESSAGE:
[ERROR] sh: /usr/omni/bin/omnitrig: not found
CAUSE:
The omnitrig is missing and crontab was not updated. This can happen
if OmniBack II was removed instead of deinstalled.
ACTION:
Manually remove the omnitrig line from the crontab file.
10-90
Glossary
This glossary contains definitions
and additional information about
words and phrases used in HP
OpenView OmniBack II.
access permissions Unix File name
characteristics (such as read, write, and
execute) which determine whether a user
can perform an operation on a file (such
as opening a file for writing). This is
different from access rights which
control which tasks an OmniBack II user
can perform.
from a single workstation, the Cell
Server.
Cell Console The user interface of
OmniBack II. It usually runs on the Cell
Server, but can also be installed on any
system in the cell.
Cell Server This is the workstation
from which a cell is managed. It includes
a Cell Console, several session
managers, all configuration files, and the
OmniBack II log database.
access rights The permission to
perform specific backup and restore
tasks, such as handling mount prompts.
Users have the access rights of the user
class to which they belong.
client A computer on a network that
receives service from another computer
(server). For example, a communication
server connects clients to other
communications networks such as local
area networks (LANs). (See server.)
ACS pool Is used to optimally manage
and control StorageTek Library systems.
complex device See Device Chains,
Exchangers, or Stackers.
cell A local part of the network. The
system administrator groups several
workstations, hosts, devices, and other
items necessary for backup into a cell to
more easily manage backup and restore
operations. A cell typically represents a
location or organizational structure.
Backup and restore policies and
schedules are defined on the cell level.
This lets system administrators
customize the backup and restore
functions for each cell. By default, a cell
does not cross LAN boundaries. Also
known as "domain".
concurrency Media multiplexing.
This device option allows more than one
Disk Agent (up to five) to write to one
backup device. This helps OmniBack II
keep the media streaming when the
device can accept data faster than a disk
can send it. OmniBack II provides
defaults for each device type supported
cell configuration Setting up hosts,
pools, and devices for a cell and defining
the services and policies for the cell.
Once configured, a cell can be controlled
control device The device file used to
control the robotic arm which moves the
media between the repository slots and
the read-write mechanism of an
condition The quality and reliability
of a medium. Heavy usage and age result
in increased read and write errors. Media
should be replaced when the condition
field indicates POOR.
Glossary-i
Glossary
exchanger.
DAS pool Is used to optimally manage
and control GRAU systems.
daemon A memory resident program
that is constantly running. The
OmniBack II daemon facilitate
communications in the network so that
OmniBack II can access different parts
of the network at any time.
database server A computer with a
large database stored on it. A server has
a database that can be accessed by client
workstations. (See also Client).
datalist A list of objects to be backed
up. These objects can be files or
directories. OmniBack II backs up the
objects in the datalist.
default entry A datalist entry that
establishes default values for a group of
objects.
default selection An object or action
that is used when no other selection is
specified.
device A unit that contains stored data
for backup or recovery. See also
Complex Device and Source Device.
detail catalog A catalog in the
OmniBack II database containing the
detail information from backup and
restore sessions. You can delete this
catalog if you run out of disk space and
you will still be able to restore all files
backed up, but it will take longer time.
Glossary-ii
Disk Agent The OmniBack II
software module that controls reading
from and writing to a disk. During a
backup session the Disk Agent sends
data from the disk to the Media Agent
which then stores the data to the media.
During a restore session the Disk Agent
takes data received from the Media
Agent and writes it to the disk.
OmniBack II only backs up those
systems which have the Disk Agent
module installed.
exchangers Also called autochangers,
jukeboxes, etc. Standalone storage
devices. An exchanger contains a library
system of media which are held in
repository slots. Each slot holds one
medium (for example, DAT tape or
magneto- optical disk). The exchanger
provides access to a vast library of data
within a few seconds without operator
intervention. To read a medium, an
electro-mechanical arm, guided by the
host computer, selects, moves, rotates,
and inserts tape or optical cartridges into
the drive mechanism(s) mounted inside
the cabinet.
exchanger data device A data storage
device. This device includes repository
slots, a read mechanism, and a robotic
arm which moves media between the
repository slots and the read mechanism.
exchangers pool Is used to optimally
manage and control large SCSI II
exchangers.
exchanger repository slots Make up
an exchanger repository. Each slot holds
one storage medium, for example, DAT
Glossary
tape or magneto-optical disk. OmniBack
II references each slot by a number. To
read a medium, a control arm removes
the medium from a slot and inserts it in
the read mechanism.
full backup A backup in which all
selected objects on all selected objects
are backed up whether or not they have
been modified recently. This contrasts
with incremental backup.
export host To remove a host from the
Cell Configuration without deinstalling
the OmniBack II modules which are on
this host. This action allows you to move
a host from one cell to another without
deinstalling and reinstalling the
OmniBack II modules. You export the
host from one cell and import the host to
another cell.
host A server with workstations
including the disks and disk drives of the
workstations.
export medium To remove all
information about a backup medium and
its contents from the OmniBack II
database of this cell. Once a medium is
exported, OmniBack II no longer has
information about either the medium
itself (for example its location) or what
backup sessions (and files) are contained
on this medium. Use this option when
the medium is no longer to be used for
backup in this cell. Data on the medium
remains unchanged. See also import
medium.
hostname The name of the systemin
the network.
import host Adds an already installed
host to the OmniBack II cell. See also
export host.
import medium Rereads all data about
OmniBack II sessions back to the
database. See also export medium.
IP Address The numeric address of a
host. The IP address consists of 4 groups
of numbers: the first three groups of
numbers represent the number of the
internet network and the last group of
numbers represents the local network
address.
filter Criteria which let you selectively
display information in OmniBack II
windows. Use filters to define a view of
session activities based on a specific
session owner or type of media session
(backup, restore). If you use a filter to
display names, OmniBack II generally
allows you to use wildcards to view a
particular pattern.
IPC address
See IP Address.
force initialization To erase any data
contained on a medium and format the
medium as an OmniBack II medium.
LAN An acronym for Local Area
Network. The systems and/or clusters
which share data, hardware, and
incremental backup A backup in
which only objects changed since a
recent full backup are copied. Related to
full backup.
label A user-defined identifier used to
describe a backup medium.
Glossary-iii
Glossary
software resources via Networking
Services software.
location The physical location of the
backup media (for example, "building 4"
or "offsite storage").
logical device A set of physical
devices treated and accessed as if it were
one device. A logical device consists of
devices of the same type (such as DAT,
optical or Exabyte) and usage policy.
media class The format of the media
(for example DAT)
media condition factors The userassigned age threshold and usage
threshold used to calculate the probable
condition of the medium. The following
terms describe the condition:
Good = indicates less than 80% of
thresholds for age or usage
magazine pool Is used to optimally
manage and control small exchanger
devices, such as HP SureStore 12000e.
Magazine pool also supports the use of
cleaning tapes.
Fair = indicates 81-100% of the
thresholds for age or usage
media Physical objects upon which
data is recorded, such as DAT tapes or
optical disks. You save or back up data to
media. The plural of medium.
Default values for the thresholds are
provided by OmniBack II.
Media Agent The OmniBack II
software module that controls reading
from and writing to a media drive.
During a backup session, the Media
Agent writes to media the data it has
received from the Disk Agent. During a
restore session, the Media Agent locates
data on the backup media and sends the
data to the Disk Agent. The Disk Agent
then writes the data to the disk.
media allocation policy determines in
which order media are accessed within a
pool. Strict allocation policy directs
OmniBack II to prompt for a specific
medium. Loose directs OmniBack II to
prompt for any suitable medium.
media capacity See medium capacity
Glossary-iv
Poor = indicates more than 100% of the
thresholds for age or usage
media ID A unique identifier assigned
by OmniBack II to a medium.
media label
See label.
media management A tool that
allows you to track your backup media
and backup/restore history. Media
management also protects your data by
helping to prevent accidental
overwriting of data.
media pool A set of like media
(DATs, Exabytes, etc.) used and tracked
as a group. Media in a pool must be the
same media class (i.e. DAT). Media are
initialized and assigned to a media pool.
See also standard pool, magazine pool,
exchanger pool, ACS pool, and DAS
Glossary
pool.
medium (See media.)
medium capacity The space available
on a medium before any data is written to
the medium. If determine is selected,
OmniBack II determines the media
capacity for that media type (for
example, DAT). If specify is selected,
OmniBack II uses the capacity that you
specified when you initialize the
medium.
merge Defines how the files are
restored to the destination. If files exist
in the destination path, those with the
newer date are kept. Not existing files
are always restored to the destination
path. See also overwrite.
mount To add an auxiliary
(removable) file system to an active
existing file system.
mount prompt A screen prompt that
tells the user to put media online. This
prompt appears in the status field of the
monitor windows. When a mount
prompt has been answered, the status
field changes to in progress.
mount point The access point to the
data in the directory structure.
node
A host in a cell.
object An entity such as a directory,
file, filesystem or database that is used in
a backup.
overwrite
Defines how files are
restored to the destination. All files are
restored from the backup even if they are
older than existing. See also merge.
permissions Unix file characteristics
(such as read, write, or execute) which
determine whether you can perform a
certain operation on the file (such as
writing to a file).
pool (See media pool)
pool attributes Attributes that define
the characteristics and use of the pool.
Different pool types have different
attributes.
pre-exec An option that instructs
OmniBack II to process a command or
script before it backs up an object or the
session is started.
post-exec An option that instructs
OmniBack II to process a command after
the backup of an object or the session is
completed.
recycle OmniBack II keeps track of
data on every used medium. Recycling
removes data protection from the data
backed up, thus allowing OmniBack II to
overwrite data in the next backup. The
data on the media remains unchanged.
repository slots See exchanger
repository slots.
rescan A logical device function that
checks each slot of the exchanger to
identify the media and update the media
management database with this
information. If someone has manually
Glossary-v
Glossary
removed or switched media, do a rescan.
Rescanning lets OmniBack scan all the
slots in the exchanger, confirm what
media is present in each slot, update and
correct any inconsistencies within the
media management database.
scheduler The function that controls
how often and when backups occur. By
setting up a schedule for a datalist, you
can automate your backups to run
periodically at times of low system
usage. With the scheduler, the backups
you set up will run on the days and at the
times you specify without your
attendance as long as the devices and
media are properly set.
server A computer on a network that
provides service to terminals on the
network (clients) by managing an
expensive shared resource. For example,
a file server manages a set of disks and
provides storage services to computers
(clients) on the network that may not
have their own disks. Sometimes a
server is a dedicated computer. (See also
client.)
session A major operation of
OmniBack II such as backup or restore.
Each session is monitored and controlled
by a Session Manager.
source device The mass storage unit
from which OmniBack II is to be
installed or updated, for example a
cartridge tape, a DDS, or a CD-ROM.
standard pool Is used to manage
standalone devices, such as DATs,
jukeboxes, or small external exchangers.
Glossary-vi
verify A function that lets you check
whether or not the data on a specified
medium is consistent or valid. This
function reads all blocks, checks the
headers, parses all Media Agent blocks
and checks every record within the
block.
Index
A
aborting
session, 6-3
adding
media to pool, 3-39
slots to autochanger pool, 3-38
slots to exchanger pool, 3-46
allocation policy, 3-19
autochanger pool
adding slots, 3-38
migrating to, 3-37
auxiliary devices, 3-25
B
backup
aborting, 6-3
choosing mode of backup, 3-6
configuring filesystem, 4-7
display of status, 4-16
existing datalist, 4-14
filesystem, 3-5
incremental, 3-6
interactive, 4-14
monitoring session, 6-3
options, 4-19
pre- and post- commands, 7-6
previewing, 4-18
protection option, 3-10
rawdisk, 3-6, 4-11
recovery failed, 7-12
restarting failed backups, 7-13
scheduled, 4-28
scheduling, 4-29
backup options
compression, 4-22
encryption, 4-22
ignore hard links, 4-24
lock files during backup, 4-24
log files to database, 4-23
pre- and post- commands, 7-6
pre/post exec, 4-24
preserve access time, 4-24
private/private, 4-23
protection, 3-10, 4-25
report level, 4-23
backup schedule,configuring, 3-9
backup schedules
changing date, 4-34
clearing, 4-33
for holidays, 4-34
periodic, 4-31
predefined, 4-31
backup session
deleting from media, 7-19
backup sessions
selecting for restore, 5-7
backup strategy
planning, 3-5
backup type, 3-5
backup, full, 3-6
barcodes
quick scan using barcodes, 3-45
Browse, 5-7
C
CDF files
detection, 7-14
cell
access planning, 3-11
planning, 3-3
planning backup strategy, 3-5
cell access
planning, 3-11
Cell Server, 1-5
moving, 7-26
recovering, 7-22
changing
view of backup objects, 5-4
Index-i
Index
cleaning tape, 3-26
pool attribute, 3-22
with magazine pool, 3-16
client
recovering, 7-21
command line interface, 1-12
compression module
customizing, 7-37
compression option, 4-22
concurrency option, 4-26
configuring
backup schedule, 3-9
data compression, 3-12
data encryption, 3-12
logical devices, 3-25, 3-27
media management, 3-13
user classes, 3-48
users, 3-48
corrupted files
restoring, 5-8
crash
preparing a mission critical example, 7-23
preparing for, 7-20
recovering Cell Server, 7-22
recovering client, 7-21
CRC Check option, 4-25
create
datalist for interactive backup, 4-17
creating
datalist, 4-5
media pools, 3-38
customizing
compression module, 7-37
encryption module, 7-44
D
data compression,configuring, 3-12
data encryption,configuring, 3-12
database
Index-ii
adding to datalist, 8-13
backing up, 8-13
copying to an ASCII file, 8-10
reading database from ASCII file, 8-10
restoring, 8-14
shrinking the file, 8-10
datalist
auto generate, 4-4
concepts, 4-3
creating, 4-5
for interactive backup, 4-17
modifying, 4-6
previewing backup, 4-18
starting backup, 4-14
deinstalling
OmniBack II host, 2-9
delete files
restore option, 5-12
deleting
session from media, 7-19
device
auxiliary, 3-25
dirty tape message, 3-26
logical devices, 3-25
primary, 3-25
rescanning, 3-35
with cleaning tape, 3-26
device options
concurrency, 4-26
CRC check, 4-25
media pool, 4-26
prealloc, 4-26
devices
physical, 3-26
dirty tape, 3-26
disabling
CDF files detection, 7-14
Disk Agent, 1-5
disk section
Index
restore from rawdisk backup, 5-17
display statistical information
restore option, 5-12
distributing
software to clients, 2-3
E
enabling
CDF files detection, 7-14
encryption module
customizing, 7-44
encryption option, 4-22
erasing medium, 3-40
error messages, 10-2
exchanger
rescanning, 3-35
exchanger pool
adding slots, 3-46
removing non-OmniBack II media, 3-45
exclude
restore option, 5-10
exporting
hosts from the cell, 2-6
media from pool, 3-43
F
failed backup
restarting, 7-13
file restore
options for rawdisk restore, 5-23
files
configuring backup, 4-7
corrupted, 5-8
locking during backup, 4-24
lost, 5-7
selecting for restore, 5-6
starting restore, 5-13
filesystem
configuring backup, 4-7
filesystem backup, 3-5
full backup, 3-6
G
generating datalists, 4-4
global options
modifying, 7-11
graphical user interface, 1-9
GUI, 1-9
H
hosts
deinstalling, 2-9
exporting, 2-6
importing, 2-5
moving between cells, 2-5
updating modules, 2-8
HP-UX 10.x
Journaled File System (JFS,VxFS), 7-15
I
ID
media, 3-23
ignore hard links option, 4-24
importing
hosts, 2-5
media to pool, 3-44
incremental backup, 3-6
levels, 3-7
initializing
media in pool, 3-39
interactive backup, 4-14
creating temporary datalist, 4-17
existing datalist, 4-14
J
JFS, 7-15
concepts, 7-15
limitations, 7-16
Index-iii
Index
restore, 7-17
template scripts, 7-17
L
label
media, 3-23
list restored objects
restore option, 5-12
locating files, 5-7
lock files
restore option, 5-12
log files option, 4-23
logical devices, 3-25
auxiliary, 3-25
configuring, 3-25, 3-27
primary, 3-25
using, 3-25
logical volume
restore from rawdisk backup, 5-17
lost files
restoring, 5-7
M
magazine pool, 3-15
cleaning tape attribute, 3-22
with cleaning tape, 3-16
match
restore option, 5-10
MC/ServiceGuard, 7-28
backing up the Secondary Cell Server, 7-32
configuring, 7-28, 7-29
prerequisites for configuration, 7-29
media
adding slots to autochanger pool, 3-38
adding slots to exchanger pool, 3-46
adding to pool, 3-39
compatibility, 3-22
creating media pool, 3-38
deleting a session, 7-19
Index-iv
erasing, 3-40
exporting from pool, 3-43
ID, 3-23
identification, 3-23
importing to a pool, 3-44
initializing media in pool, 3-39
label, 3-23
magazine pool, 3-15
magazine pool with cleaning tape, 3-16
media management tasks, 3-37
media pool, 3-13
migrating to the autochanger pool, 3-37
modifying, 3-41
moving, 3-42
normal pool, 3-14
physical repository pool, 3-17
recycling, 3-42
removing non_OmniBack from the exchanger’s pool, 3-45
removing protection, 3-42
scanning, 3-45
verifying, 3-40
viewing information about, 3-41
viewing scheduled for backup, 3-47
Media Agent, 1-5
media allocation policy, 3-19
media class, 3-19
media condition factors, 3-21
media management, 3-13
media pool, 3-13
adding media, 3-39
creating, 3-38
device option, 4-26
initializing media, 3-39
magazine pool, 3-15
migrating to autochanger pool, 3-37
normal pool, 3-14
physical repository pool, 3-17
removing non-OmniBack II media, 3-45
Index
media pools
pool name, 3-19
media usage policy, 3-20
merge
restore option, 5-11
messages, 10-2
messages field
status description, 4-16
mission critical system, 7-23
mode of backup, 3-6
modify
datalist, 4-6
modifying
datalist, 4-6
global options, 7-11
media, 3-41
monitoring
completed sessions, 6-5
running sessions, 6-3
several cells, 6-7
viewing details of a running session, 6-3
mount prompt
automatic handling, 7-3
responding, 5-20
move busy files
restore option, 5-12
moving
cell server, 7-26
medium, 3-42
N
network load, 5-14
no overwrite
restore option, 5-11
normal pool, 3-14
O
OmniBack II
architecture, 1-4
backing up a database, 8-13
backup options, 4-19
backup session, 1-6
CDF files, 7-14
commands, 1-12
components, 1-4
copying database to ASCII file, 8-10
datalist, 4-3
deinstalling hosts, 2-9
deleting a session from media, 7-19
distributing to clients, 2-3
error messages, 10-2
exporting hosts, 2-6
filesystem backup, 4-7
generating reports, 7-33, 7-34, 7-35
global options file, 7-11
graphical user interface, 1-9
HP OperationsCenter configuring, 3-11
importing hosts, 2-5
interactive backups, 4-14
MC/ServiceGuard, 7-28
configuring, 7-28, 7-29
prerequisites for configuration, 7-29
planning backup strategy, 3-5
rawdisk backup, 4-11
reading database from ASCII file, 8-10
restore session, 1-7
restoring database, 8-14
scheduling backup, 4-29
shrinking the database, 8-10
software modules, 2-3
updating modules, 2-8
with JFS (VxFS) on HP-UX 10.0, 7-15
OmniBack II database
limiting growth, 8-6
moving to another system, 8-17
projecting growth, 8-4
omninotify.ksh command, 7-34
Index-v
Index
omnirpt.ksh script, 7-35
options
for backup, 4-19
pre- and post- commands, 7-6
protection, 3-10
restore, 5-9
overwrite
restore option, 5-11
ownership
changing, 4-26
P
physical devices, 3-26
physical repository pool, 3-17
planning
cell, 3-3
cell access, 3-11
pool
creating, 3-38
exporting media, 3-43
importing media to, 3-44
magazine pool, 3-15
media class attribute, 3-19
normal, 3-14
physical repository, 3-17
pool attributes
cleaning tape, 3-22
media allocation policy, 3-19
media class, 3-19
media condition factors, 3-21
media usage policy, 3-20
pool name, 3-19
pool name, 3-19
post- exec
restore option, 5-13
post- exec commands, 7-6
post-exec command, 4-24
pre- exec
restore option, 5-12
Index-vi
pre- exec commands, 7-6
prealloc option, 4-26
pre-exec command, 4-24
preserve access time attributes, 4-24
previewing backups, 4-18
previous session
viewing, 6-5
primary logical device, 3-25
private option, 4-23
protection
removing, 3-42
protection attributes
restore option, 5-11
protection option, 3-10, 4-25
public option, 4-23
R
rawdisk
configuring backup, 4-11
rawdisk backup, 3-6
rawdisk restore
files, 5-21
options for file restore, 5-23
section, logical volume, 5-17
recanning a device, 3-35
recovering, 7-20
Cell Server, 7-22
client system, 7-21
failed backups, 7-12
recovery
mission critical system example, 7-23
recycling media, 3-42
removing
non-OmniBack II media from the exchanger
pool, 3-45
protection, 3-42
report level option, 4-23
reports
generating, 7-33
Index
omninotify.ksh command, 7-34
omnirpt.ksh script, 7-35
repository, 9-7, 9-13
respond
to mount prompt request, 5-20
restarting
failed backups, 7-13
restore
changing the view of backup objects, 5-4
files from a rawdisk backup, 5-21
monitoring sessions, 6-3
network load, 5-14
options for file restore from rawdisk, 5-23
rawdisk section, logical volume, 5-17
responding to mount prompt, 5-20
restore as option, 5-9
searching for files, 5-15
selecting files, 5-6
starting, 5-13
target hostname option, 5-11
restore into
restore option, 5-10
restore options
delete files, 5-12
display statistical information, 5-12
exclude, 5-10
for fileset, 5-9
list restored objects, 5-12
lock files, 5-12
match, 5-10
merge, 5-11
move busy files, 5-12
no overwrite, 5-11
overwrite, 5-11
post- exec, 5-13
pre- exec, 5-12
protection attributes, 5-11
restore as, 5-9
restore into, 5-10
skip, 5-10
sparse files, 5-12
target hostname, 5-11
time attributes, 5-11
restoring
corrupted files, 5-8
lost files, 5-7
overview, 5-3
S
scan
quick scan using barcodes, 3-45
scanning
media, 3-45
scheduling
backups, 4-28
searching for files, 5-15
selecting
files to restore, 5-6
selecting backup objects, 5-5
session
aborting, 6-3
deleting from media, 7-19
monitoring, 6-3
previewing backup, 4-18
status, 4-16
viewing completed, 6-5
viewing details, 6-3
viewing details of a completed session, 6-6
Session Manager, 1-5
skip
restore option, 5-10
slots
adding to autochanger pool, 3-38
adding to pool, 3-46
sparse files
restore option, 5-12
starting
failed backups, 7-13
Index-vii
Index
graphical user interface, 1-10
restore, 5-13
status
color description, 4-16
messages description, 4-16
switch ownership, 4-26
system
recovering after crash, 7-20
T
tasks
media management, 3-37
time attributes
restore option, 5-11
troubleshooting, 10-2
U
updating
OmniBack II modules, 2-8
usage policy
pool attribute, 3-20
user classes
configuring, 3-48
users
configuring, 3-48
using logical devices, 3-25
using media management, 3-37
V
verifying
medium in pool, 3-40
viewing
information about media, 3-41
media scheduled for backup, 3-47
VxFS
with OmniBack II, 7-15
Index-viii
Index
Index-ix
Open as PDF
Similar pages