IBM Tivoli Storage Manager for System Backup and Recovery

IBM Tivoli Storage Manager for System Backup and
Recovery
Installation and User’s Guide
Version 5 Release 5
GC32-9076-00
IBM Tivoli Storage Manager for System Backup and
Recovery
Installation and User’s Guide
Version 5 Release 5
GC32-9076-00
First Edition (December 2002)
This edition notice applies to the IBM Tivoli Storage Manager for System Backup and Recovery Version 5 Release 5
(program number 5698-SYS) and to any subsequent releases until otherwise indicate in new editions or technical newsletters.
IBM Tivoli Storage Manager for System Backup and Recovery will be referred to as “SysBack” in certain text within
this document.
The following paragraph does not apply to the United Kingdom or any country where such provisions are
inconsistent with local law: INTERNATIONAL BUSINESS MACHINES CORPORATION PROVIDES THIS
MANUAL “AS IS” WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING,
BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A
PARTICULAR PURPOSE. Some states do not allow disclaimer of expressed or implied warranties in certain
transactions; therefore, this statement may not apply to you.
This publication could include technical inaccuracies or typographical errors. Changes are periodically made to the
information herein; these changes will be incorporated in new editions of this publication. IBM may make
improvements and/or changes in the product(s) and/or program(s) described in this publication at any time.
It is possible that this publication may contain references to, or information about, IBM products (machines and
programs), programming, or services that are not announced in your country. Such references or information must
not be construed to mean that IBM intends to announce such IBM products, programming, or services in your
country. Any reference to an IBM licensed program in this publication is not intended to state or imply that you can
use only IBM’s licensed program. You can use any functionally equivalent program instead.
Requests for copies of this publication and for technical information about IBM products should be made to your
sales office or the IBM Branch Office serving your locality.
IBM and AIX are registered trademarks of International Business Machines Corporation.
û Copyright International Business Machines Corporation 2002. All rights reserved.
US Government Users Restricted Rights - Use, duplication or disclosure restricted by GSA ADP Schedule Contract
with IBM Corp.
© Copyright International Business Machines Corporation 2002. All rights reserved.
US Government Users Restricted Rights – Use, duplication or disclosure restricted by GSA ADP Schedule Contract
with IBM Corp.
Contents
IBM Tivoli Storage Manager for System Backup and Recovery Overview . . . . . . . ix
About This Book. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi
Who Should Read This Publication . . . .
IBM Tivoli Storage Manager for System Backup
Conventions Used in This Book . . . . . .
Contacting Customer Support . . . . . .
Trademarks . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . xi
and Recovery Documentation . . . . . . . . . . . . . xi
. . . . . . . . . . . . . . . . . . . . . . . xi
. . . . . . . . . . . . . . . . . . . . . . . xi
. . . . . . . . . . . . . . . . . . . . . . . xii
Chapter 1. Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-1
Prerequisite Tasks . . . . . . . . . . . .
Upgrading from SysBack Backup & Recovery for AIX
Procedure for Installation from CDROM . . . .
Accessing the Online Documentation . . . . .
. . . .
- SysBack
. . . .
. . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1-1
1-1
1-1
1-3
Chapter 2. SMIT Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-1
Accessing the IBM Tivoli Storage Manager for System Backup
How the SysBack Menus are Organized . . . . . . .
Backup & Recovery Options . . . . . . . . . .
Configuration Options . . . . . . . . . . . .
Tape Drives . . . . . . . . . . . . . . . .
Utilities . . . . . . . . . . . . . . . . .
Device Selector Screens . . . . . . . . . . . . .
Backups to Disk Image Files . . . . . . . . . . .
Backup, List, Verify and Restore Command Output Screen .
SMIT Help Screens . . . . . . . . . . . . . .
SMIT Fastpaths . . . . . . . . . . . . . . . .
Obtaining Command Line Options from SMIT . . . . .
and Recovery
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
2-1
2-2
2-2
2-2
2-3
2-4
2-4
2-6
2-7
2-9
2-9
2-9
Chapter 3. License Configuration . . . . . . . . . . . . . . . . . . . . . . . . 3-1
Chapter 4. Performing Backups . . . . . . . . . . . . . . . . . . . . . . . . . 4-1
Understanding Incremental Backups . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-1
Understanding Pull Backups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-3
Backing up the System (Installation Image) . . . . . . . . . . . . . . . . . . . . . . . . 4-3
Backing up Volume Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-9
Backing up Filesystems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-14
Backing up Logical Volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-19
Backing up Files or Directories . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-24
Chapter 5. Listing or Verifying Backups . . . . . . . . . . . . . . . . . . . . . 5-1
Listing Backup Contents . . . .
Verifying Files on System Backup .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
. 5-1
. 5-2
Chapter 6. Recreating or Restoring from Backups . . . . . . . . . . . . . . . . . 6-1
Removing Volume Groups, Logical Volumes, and Filesystems
Recreating Volume Groups, Logical Volumes, and Filesystems
Understanding Incremental Restores . . . . . . . . .
Restoring Data from a Backup . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
6-1
6-2
6-6
6-7
Chapter 7. Scheduled Backups and Scripts. . . . . . . . . . . . . . . . . . . . 7-1
Scripts . . . . . . . . . . . .
Accessing the Backup Scripts Menu .
© Copyright IBM Corp. 2002
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
. 7-1
. 7-1
iii
Creating a Backup Script File . . .
Changing a Backup Script File . . .
Removing a Backup Script File. . .
Accessing the Backup Schedules Menus
Creating a Backup Schedule . . .
Updating a Backup Schedule . . .
Listing Backup Schedules . . . .
Removing a Backup Schedule . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
7-1
7-3
7-4
7-4
7-4
7-6
7-6
7-7
Chapter 8. Local User Access . . . . . . . . . . . . . . . . . . . . . . . . . 8-1
Understanding Local User Access Permissions . . . .
Adding or Changing Local User Device/Directory Access
Listing User Access . . . . . . . . . . . . .
Removing User Access . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
8-1
8-2
8-4
8-4
Chapter 9. Remote Services . . . . . . . . . . . . . . . . . . . . . . . . . . 9-1
Understanding Remote Host and User Access Permissions . .
Backing Up, Listing, Verifying, or Restoring from Remote Backup
Configure Remote Services . . . . . . . . . . . . .
Adding or Changing Client Host Access to This Server . . .
Listing Client Access . . . . . . . . . . . . . . .
Removing Client Access . . . . . . . . . . . . . .
Adding or Changing the List of Remote Backup Servers . . .
Listing Remote Backup Servers . . . . . . . . . . .
Removing a Remote Backup Server . . . . . . . . . .
Remote Commands Access for use with Pull Backups . . . .
. . .
Devices
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
9-2
9-4
9-4
9-4
9-7
9-7
9-7
9-9
9-9
9-9
Chapter 10. Virtual Devices . . . . . . . . . . . . . . . . . . . . . . . . . . 10-1
Types of Virtual Devices .
Volume Numbering . . .
Autoloaders . . . . .
Adding a Virtual Device .
Listing Virtual Devices . .
Removing a Virtual Device
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
10-2
10-2
10-3
10-3
10-5
10-5
Chapter 11. Activity Logging. . . . . . . . . . . . . . . . . . . . . . . . . . 11-1
Accessing the SysBack Logging Menus . . . . . . . .
Changing the SysBack Logging Attributes (Size and Location) .
Displaying the SysBack Log . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
. 11-1
. 11-1
. 11-2
Chapter 12. Exclude Lists . . . . . . . . . . . . . . . . . . . . . . . . . . . 12-1
Adding a File, Directory or Logical Volume to an Exclude List.
Listing Excluded Files or Directories . . . . . . . . .
Removing Files or Directories from an Exclude List . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
. 12-2
. 12-3
. 12-4
Chapter 13. Network Boot/Installation Configuration . . . . . . . . . . . . . . . 13-1
Classic Network Boot and NIM Resource Boot . . . . . . . . . . . . . . . . . . . . . . . 13-1
Classic Boot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13-1
NIM Resource Boot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13-2
Classic Network Boot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13-2
Accessing and Configuring the Classic Network Boot Menus . . . . . . . . . . . . . . . . . 13-2
Configuring or Updating a Network Boot Image . . . . . . . . . . . . . . . . . . . . . 13-3
Adding or Changing a Network Boot Client . . . . . . . . . . . . . . . . . . . . . . . 13-3
Setting the Network Install Client Defaults . . . . . . . . . . . . . . . . . . . . . . . 13-5
Rebuilding Network Boot Images . . . . . . . . . . . . . . . . . . . . . . . . . . 13-9
Removing a Network Boot/Install Client . . . . . . . . . . . . . . . . . . . . . . . . 13-9
NIM Resource Boot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13-9
Accessing and Configuring the NIM Resource Menus . . . . . . . . . . . . . . . . . . . 13-9
Installing SysBack into a SPOT Resource. . . . . . . . . . . . . . . . . . . . . . . . 13-10
iv
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Querying for SysBack Installation in SPOT Resource.
Adding or Changing a Network boot Client . . .
Rebuilding Network Boot Images . . . . . . .
Removing a Network Boot/Install Client . . . .
Initiating an SP Node Boot and Install . . . . .
Chapter 14. Tape Drives
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
13-11
13-11
13-12
13-13
13-13
. . . . . . . . . . . . . . . . . . . . . . . . . . . 14-1
Listing tape drives . . . . . . . . .
Configuring a defined tape drive . . . .
Unconfiguring an Available Tape Drive . .
Changing Tape Drive Characteristics . . .
Unloading or Ejecting a Tape from the Drive
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
14-1
14-2
14-2
14-2
14-3
Chapter 15. Utilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15-1
Displaying SysBack Product Installation History . . .
Printing or Displaying System Configuration . . . .
Creating a Bootable Tape (Without Data) . . . . . .
Changing SysBack Tape Boot Defaults . . . . . . .
Changing the Read Permission on a Disk Image File . .
Creating or Editing an LVM Information File . . . .
Creating a Custom Installation Diskette . . . . . .
Changing SysBack Program Defaults . . . . . . .
Backup Format Settings . . . . . . . . . .
Error Handling Settings . . . . . . . . . .
SysBack Network Settings . . . . . . . . .
Enabling Large File Support after Migration to AIX 4.2 .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
15-1
15-2
15-3
15-5
15-7
15-9
15-11
15-13
15-14
15-15
15-16
15-17
Chapter 16. System Installation and Maintenance . . . . . . . . . . . . . . . . 16-1
Recovery Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-1
No-prompt Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-1
Handling Installation Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-2
The SysBack Installation and Maintenance Menu . . . . . . . . . . . . . . . . . . . . . . 16-2
Changing the Installation Device . . . . . . . . . . . . . . . . . . . . . . . . . . 16-3
Changing Volume Group & Logical Volume Information. . . . . . . . . . . . . . . . . . . 16-6
Installing the System with Current Settings . . . . . . . . . . . . . . . . . . . . . . . 16-8
Utilities Menu . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-10
Rebooting the System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16-15
Chapter 17. RS/6000 Scalable POWERParallel (SP) Boot and Install Utilities . . . . . 17-1
Advantages of the SysBack SP Boot and Installation Utilities
How SysBack works with NIM on the SP System . . . .
Special Assumptions . . . . . . . . . . . . . .
Accessing the SP Boot and Install Options . . . . . .
Completing the Initiate SP Node Boot and Install Menu . .
Unprompted and Prompted Installations . . . . . . .
Troubleshooting Tips . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Chapter 18. Changing the Volume Group, Logical Volume and Filesystem Attributes
.
.
.
.
.
.
.
.
.
.
.
.
.
.
17-1
17-2
17-2
17-3
17-3
17-4
17-4
18-1
Selecting Physical Volumes for Volume Groups . . . . . . . . . . . . . . . . . . . . . . . 18-2
Changing Volume Group Attributes . . . . . . . . . . . . . . . . . . . . . . . . . . 18-4
Selecting Physical Volumes for Logical Volumes . . . . . . . . . . . . . . . . . . . . . . 18-5
Changing Logical Volume Attributes . . . . . . . . . . . . . . . . . . . . . . . . . . 18-7
Changing Filesystem Attributes. . . . . . . . . . . . . . . . . . . . . . . . . . . . 18-12
Appendix A. Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-1
cfgremrootaccess . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-2
chimagefile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-4
chinstclient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-6
cfglocaccess . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A-10
Contents
v
cfgremaccess . . . .
cfgremserver . . . .
cfgremsvs . . . . .
cfgvirdev . . . . .
chexcludelist . . . .
editlvminfo . . . .
getlvminfo . . . . .
lscfginfo. . . . . .
mkdirback . . . . .
mkjfsback . . . . .
mklvback . . . . .
mksbnetboot . . . .
mkvgback . . . . .
mkvginfo . . . . .
readsbheader . . . .
remakevg . . . . .
sbalog . . . . . .
sbclient . . . . . .
sbcomp and sbuncomp.
sbcron . . . . . .
sbdevice . . . . .
sbejecttape . . . . .
sbeot . . . . . . .
sbfwd . . . . . .
sbread . . . . . .
sbscript . . . . . .
sbspboot . . . . .
sbspotcust . . . . .
sbwrite . . . . . .
sysback . . . . . .
sysrestore . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
A-12
A-15
A-17
A-18
A-20
A-22
A-23
A-27
A-29
A-33
A-37
A-41
A-45
A-49
A-53
A-55
A-58
A-59
A-62
A-63
A-65
A-67
A-68
A-69
A-71
A-74
A-76
A-77
A-79
A-84
A-90
Appendix B. Booting a System for SysBack Installation or Maintenance . . . . . . . B-1
Initiating the boot process . . . . . . . . . . . . . .
Common Hardware Reference Platform (uni or multi-processor)
Microchannel-based RISC System/6000 (uniprocessor) . . . .
Microchannel-based RISC System/6000 SMP (multiprocessor) .
PCI-based (RSPC) RISC System/6000 (uni or multi-processor) .
Initiating the Network Boot. . . . . . . . . . . . . .
Completing the Boot Process . . . . . . . . . . . . .
Troubleshooting a Network Boot . . . . . . . . . . . .
Determining the Network Adapter Hardware Address . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
B-1
B-1
B-1
B-2
B-3
B-4
B-4
B-5
B-5
Appendix C. Creating Scripts for Customizing the System Backup and Install Process C-1
Script Names . . . .
Post-Installation Scripts .
Sample Scripts . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
. C-1
. C-3
. C-3
Appendix D. Device/System-Specific Information . . . . . . . . . . . . . . . . . D-1
IBM 7208 8mm Tape Drives . . . . . . . . . .
IBM 3490, 3570, 3590, 7205 and 7337 Tape Drives . . .
Other Tape Drives. . . . . . . . . . . . . .
IBM 7331 8mm Tape Library . . . . . . . . . .
IBM 7332 4mm Tape Library . . . . . . . . . .
Other Tape Libraries or Autoloaders . . . . . . .
IBM 7133 Serial Storage Architecture (SSA) disk subsystem
IBM 7135 RAIDiant Array . . . . . . . . . . .
IBM 7137 and Other RAID Devices . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
D-1
D-1
D-1
D-2
D-2
D-2
D-3
D-3
D-3
Appendix E. LED Codes . . . . . . . . . . . . . . . . . . . . . . . . . . . . E-1
vi
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . X-1
Contents
vii
viii
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
IBM Tivoli Storage Manager for System Backup and Recovery
Overview
IBM Tivoli Storage Manager for System Backup and Recovery (also known as
“SysBack”) provides system administrators and other system users with a simple,
efficient way to backup and recover data from a command line or a SMIT
menu-driven interface. SysBack lets you recover all or part of the system. SysBack
is also flexible; you can install one system installation image to another system
with either identical or different hardware configurations, called “cloning”
SysBack features let you:
v Create various types of backups, including:
– Full system (installation image)
– Volume groups
– Logical volumes (raw data)
– Filesystems
– Specific directories or files
v Incrementally backup and restore data.
v Perform “power” system backups that enable faster backup and restore times by
backing up all data as raw logical volumes and heighten read/write
performance while reducing CPU usage.
v Perform backups to locally attached tape drives or files on disk and remote
hosts across the network.
v Selectively exclude specific files, directories, filesystems, or logical volumes from
backups.
v Centrally manage backup clients using “pull” backups from a single server.
v Create backup scripts for easy automation.
v Define backup schedules for easy automation.
v Execute pre and post-backup scripts that enable environment-specific task
automation, including halting database applications before beginning a backup.
v Perform backups to multiple sequential devices, automatically continuing the
backup on the next device when the first is full and minimizing manual
intervention when autoloading libraries are not available.
v Perform backups to multiple parallel devices, called “striping,” which lets you
complete a single backup in a fraction of the normal time.
v Create multiple copies of a single backup to different devices in approximately
the same time it takes for a single copy.
v View progress status indicators that display estimated backup or restore sizes,
times, performance estimates and a completion percentage estimate.
v Receive completion status logs on all backup, list and verification operations.
v Use SMIT menus to configure SysBack options, which let you backup and
restore volume groups, logical volumes, filesystems, directories, or files and list
and verifying backup images.
v Use sequential autoloading devices to minimize manual intervention and tape
loading operations.
v “Stack” multiple backups on a single tape for all backup types.
© Copyright IBM Corp. 2002
ix
v Reinstall the system to its original device configuration, including the volume
group and logical volume placement on disk and attached devices, using Full
System (Installation) Image to the same or different system (also called
“cloning”).
v Install the system from local boot/installation media.
v Perform a network boot or installation to alleviate the need for local boot or
installation media using SysBack functions or existing Network Installation
Manager (NIM) resources.
v Perform Recovery Installation that restores only the base operating system (root
and /usr filesystems) without affecting other data in the rootvg or other volume
groups.
v Preserve multi-copied (mirrored) or striped logical volumes on recreate.
v Import, ignore, or recreate/restore each volume group or logical volume from a
single backup media during installation.
v Retain exact partition placement of logical volumes or make contiguous
partitions of any that have become fragmented; this will improve I/O
performance on recreate.
v Execute post-installation scripts to perform environment specific tasks.
v Use post-installation to remove SysBack or network configuration after cloning a
backup image to a different machine or to indicate the location of installation
media for automatically installing device support to the new machine.
v Set network and tape boot defaults to minimize, and often eliminate, the need
for user prompting during an installation.
v Edit or change the physical location and attributes of volume groups, logical
volumes, and filesystems during recreation.
v Restore data at the volume group, logical volume, filesystem, directory or file
level.
v Interactively select files to restore by letting you specify a filelist, use a search
word or use a wildcard to narrow the list.
v Exclude select filesystems or logical volumes during restore operations.
x
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
About This Book
This book contains information on configuring and using IBM Tivoli Storage
Manager for System Backup and Recovery (also known as “SysBack”). The book
contains instructions on using the System Management Interface Tool (SMIT)
menus for performing backups as well as a commands reference to perform
SysBack functions without the SMIT menus.
Who Should Read This Publication
This book is intended for system administrators responsible for managing the AIX
operating system on RS/6000® or pSeries™ systems. Readers of this book are
expected to have a basic understanding of the SMIT menus and have already
configured devices for backing up the operating system. Readers should also be
familiar with the general layout of the AIX system, including an understanding of
the volume groups, logical volumes, and filesystems.
IBM Tivoli Storage Manager for System Backup and Recovery
Documentation
IBM posts publications for this and all other Tivoli products, as they become
available and whenever they are updated, to the Tivoli Information Center Web
site. You can access updated publications in the Tivoli Information Center from the
following Customer Support for Tivoli products Web site:
http://www.tivoli.com/support/documents/
Conventions Used in This Book
The following highlighting conventions are used in this book:
Bold
Italics
Monospace
Identifies command, keywords, files, directories, and other items
whose names are predefined by the system.
Used for emphasis and to identify parameters whose actual names
or values are to be supplied by the user.
Identifies examples of specific data values, text similar to what you
might see displayed, messages from the system, or information
you should actually type.
Contacting Customer Support
For support of this or any Tivoli product, you can contact Tivoli Customer Support
in one of the following ways:
v Visit the Storage Manager technical support Web Site at:
http://www.ibm.com/software/sysmgt/products/support/
v Submit a problem management report record (PMR) electronically at
IBMSERV/IBMLINK. You can access IBMLINK at:
http://www2.ibmlink.ibm.com
v Submit a problem management record (PMR) electronically at:
http://www.ibm.com/software/support/probsub.html
v Customers in the United States can also call 1-800-IBM-SERV.
© Copyright IBM Corp. 2002
xi
v International customers should consult the web site for customer support
telephone numbers.
v You can also review the IBM Software Support Guide which is available on our
web site at: http://techsupport.services.ibm.com/guides/handbook.html
Trademarks
The following terms are trademarks of International Business Machines
Corporation in the United States or other countries or both:
v AIX
v IBM
v IBMLink
v Magstar
v POWERparallel
v
v
v
v
v
RISC System/6000
RS/6000
SP
SysBack
Tivoli
Other company, product, and service names may be trademarks or service marks
of others.
xii
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Chapter 1. Installation
To install IBM Tivoli Storage Manager for System Backup and Recovery (SysBack),
you must have the prerequisite software installed on your system. Then, follow the
installation procedures below to install SysBack from the SysBack installation CD.
Prerequisite Tasks
The following software is required to use SysBack.
v AIX Base Operating System (BOS) Runtime Version 4.3 or higher.
v The bos.sysmgt.sysbr fileset.
v If you will be using the Remote Services functions of SysBack, install the
following:
– bos.rte.net
– bos.net.tcp.client
v If you will be using the Classic Network Boot functions for the Network Installs,
install bos.net.nfs.client
v If you will be using the NIM Resource Network Boot functions for Network
Installs, the NIM environment and resources must first be installed and
configured in the desired locations for SysBack to use these resources. For
information on installing and configuring NIM, see the book, Network Installation
Management Guide and Reference - Pub# SC23-4113.
v When AIX Version 4 is installed, the following are installed automatically with
the system and may not be removed:
– bos.rte.bosinst
– bos.rte.archive
– bos.rte.libnetsvc (when using Network Install functions)
Upgrading from SysBack Backup & Recovery for AIX - SysBack
If you are upgrading from SysBack Backup & Recovery for AIX - SysBack versions
5.1.x, 5.2.x, 5.3.x, or 5.4.x, you will not need to uninstall it before installing the IBM
Tivoli Storage Manager for System Backup and Recovery (SysBack) product.
The installation process for IBM Tivoli Storage Manager for System Backup and
Recovery will preserve any previous configurations, as well as, to automatically
make the sysback.rte fileset obsolete.
Procedure for Installation from CDROM
After the prerequisite software is installed:
1. Log in as a root user. You will see the following:
IBM AIX Operating System
(c) Copyright IBM Corp. 19XX, 19XX
(/dev/console)
login: root
2. Insert the IBM Tivoli Storage Manager for System Backup and Recovery
installation CD into the CDROM.
3. Type the AIX command smit install.
© Copyright IBM Corp. 2002
1-1
Note: This command invokes the System Management Interface Tool (SMIT),
which presents a menu-driven environment for the installation process.
The argument install is a fastpath that takes you directly to the
software installation process. The menu selections differ from one
version and release of the operating system to another, so menu
selections displayed on your system may differ slightly.
4. From the Software Installation and Maintenance menu, select Install and
Update Software.
5. Select Install and Update from the LATEST Available Software.
6. At the INPUT device/directory for software prompt, type the device name of
the CD-ROM drive. For example, /dev/cd0.
7. The following screen is displayed:
Install and Update from LATEST Available Software
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
* INPUT device/directory for software
* SOFTWARE to install
PREVIEW only? (install operation will NOT occur)
COMMIT software updates?
SAVE replaced files?
AUTOMATICALLY install requisite software?
EXTEND file systems if space needed?
OVERWRITE same or newer versions?
VERIFY install and check file sizes?
Include corresponding LANGUAGE filesets?
DETAILED output?
Process multiple volumes?
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
F3=Cancel
F7=Edit
Enter=Do
[Entry Fields]
/dev/cd0
[_all_latest]
no
yes
no
yes
yes
no
no
yes
no
yes
+
+
+
+
+
+
+
+
+
+
+
F4=List
F8=Image
Figure 1-1. The Install and Update from Latest Available Software Menu
8. To install the SysBack product, accept the default settings.
To install individual filesets, highlight Software to install and press F4.
Individually, select the filesets to install.
Note: The SysBack product fileset is tivoli.tsm.client.sysback.rte
The SysBack license filesets are tivoli.tsm.client.sysback.license.rte and
tivoli.tsm.client.sysback.license.cert
9. Press Enter to confirm selections and install the software.
10. On AIX 5.1 systems, you will be prompted to “accept”the electronic license
agreement. Indicating a “No”response will prohibit the installation of these
filesets.
1-2
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Accessing the Online Documentation
IBM posts publications for this and all other Tivoli products, as they become
available and whenever they are updated, to the Tivoli Information Center Web
Site. You can access updated publications in the Tivoli Information Center from the
following Customer Support for Tivoli products web site:
http://www.tivoli.com/support/documents/
Chapter 1. Installation
1-3
1-4
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Chapter 2. SMIT Overview
This chapter provides information about the SMIT menu interface. SMIT is used
for virtually all AIX system administrative tasks.
Accessing the IBM Tivoli Storage Manager for System Backup and
Recovery
To access the System Backup & Recovery for AIX menus, type smit to access the
System Management Interface Tool. After SysBack is installed, IBM Tivoli Storage
Manager for System Backup and Recovery is displayed at the bottom of the main
menu. Select this option to access additional SysBack screens. You can alternatively
type smit sysback to jump directly to the SysBack screens. The following is an
example of the SysBack Main SMIT Menu as shown using the ASCII SMIT menu.
IBM Tivoli Storage Manager for System Backup and Recovery
Move cursor to desired item and press Enter.
Backup & Recovery Options
Configuration Options
Tape Drives
Utilities
Backup Schedules and Scripts
F1=Help
F9=Shell
F2=Refresh
F10=Exit
F3=Cancel
Enter=Do
F8=Image
Figure 2-1. ASCII version of the SysBack Main SMIT Menu
The ASCII SMIT menu is displayed by default if AIXwindows is not running. If
AIXwindows is running, and SMIT is started from the desktop or from an aixterm
window, the graphical version of SMIT (msmit) is displayed by default. The
remainder of the screens displayed in this section use the ASCII SMIT format.
Note: To use most of the options within the SysBack menus, log in to the system
as a root user. This is for several reasons:
1. Only a root user has authority to read, and therefore backup or restore,
many files in the system.
2. When restoring files under a non-root user, that user, if permitted to read
the files, becomes the new owner of the files. Files are returned to their
original ownership only when restored by a root user.
3. Many options, such as changing the block size of the tape drive, creating
network boot images, or configuring network server and client options,
require that you have root privileges to perform the task.
© Copyright IBM Corp. 2002
2-1
How the SysBack Menus are Organized
The menus are organized by the type of backups that you can perform, followed
by additional options used for configuring or customizing additional SysBack
options. The SysBack Main Menu includes the following options:
Backup & Recovery Options
Use this option to display additional options for backing up, listing, verifying, or
restoring files from a backup depending on the type of backup performed. It is not
necessary to know the backup type to perform any of the list, verify, recreate or
restore options, as the system will read the backup to determine the backup type,
and presents only additional options that are relevant to that type of backup.
Backup & Recovery Options
Move cursor to desired item and press Enter.
Backup Options
List Backup Contents
Verify a Backup
Recreate Volume Groups, Logical Volumes & Filesystems
Restore Data from a Backup
Initiate SP Node Boot and Install
F1=Help
F9=Shell
F2=Refresh
F10=Exit
F3=Cancel
Enter=Do
F8=Image
Figure 2-2. Backup & Recovery Options
Use Backup Options to create a backup of the system that can be used to reinstall
the system from a system failure, or to use the backup to install other systems.
Additional options available from the Backup Options menu enable you to back up
a smaller part of the system or only specific files or directories.
Configuration Options
This menu provides additional options for configuration tasks, such as providing
access to local and remote users, adding and deleting entries from exclude lists,
and creating or deleting virtual device descriptions.
2-2
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Configuration Options
Move cursor to desired item and press Enter.
Local User Access
Remote Services
Remote Command Access
Virtual Devices
Exclude Lists
Network Boot/Install Configuration
License Configuration
F1=Help
F9=Shell
F2=Refresh
F10=Exit
F3=Cancel
Enter=Do
F8=Image
Figure 2-3. Configuration Options
The following options are available on this menu:
Local User Access
Provides backup permission to specific or all users on the system.
You can also indicate which local tape drives or virtual devices the
users are allowed to access and the directories to which they are
allowed to save backup images.
Remote Services
Provides backup access from remote systems to local devices or
directories. You can indicate which users on the remote hosts are
allowed access and the specific devices or directories to which they
are allowed to backup.
This option is also used on the client system to indicate which
servers to which they want to backup. Only servers that have
granted access to the client can be configured on the client.
Virtual Devices
Adds, lists, and removes virtual devices. Virtual devices are logical
device names assigned to one or more physical devices, providing
a simpler way to back up to multiple devices or autoloaders.
Exclude Lists
Enters, lists, or deletes entries from an exclude file. The exclude file
contains names of individual files or directories to exclude from all
backups generated by SysBack.
Network Boot Configuration
Creates network boot images that enable client machines to be
booted from the network, therefore requiring no local boot media.
Tape Drives
Use this option to change the physical block size of tape drives or enable or
disable tape drives on the system. This is particularly useful when you want to
temporarily disable client machines from accessing the tape drives. An additional
option enables you to unload or eject a tape cartridge from a tape drive or tape
autoloader.
Chapter 2. SMIT Overview
2-3
Tape Drives
Move cursor to desired item and press Enter.
List All Defined Tape Drives
Configure a Defined Tape Drive
Unconfigure an Available Tape Drive
Change Tape Drive Characteristics
Unload/Eject Tape from Drive
F1=Help
F9=Shell
F2=Refresh
F10=Exit
F3=Cancel
Enter=Do
F8=Image
Figure 2-4. Tape Drives
Utilities
This menu provides a list of miscellaneous options. The options on this menu
might differ depending on your level of AIX installed and might change as new
options are added to SysBack.
Utilities
Move cursor to desired item and press Enter.
Display Sysback Product Install History
Print or Display System Configuration
Create a Bootable Tape (without data)
Change Sysback Tape Boot Defaults
Change Read Permission on a Disk Image File
Create or Edit an LVM Information File
Create a Custom Installation Diskette
Change Sysback Program Defaults
Sysback Logging
F1=Help
F9=Shell
F2=Refresh
F10=Exit
F3=Cancel
Enter=Do
F8=Image
Figure 2-5. Utilities
Device Selector Screens
When selecting a backup, list, verify, recreate or restore option from any SMIT
menu, you are prompted for the device, directory, or filename to use.
Various tape devices, virtual devices, or filename options are displayed, depending
on the backup permission granted the specific user or host. When you back up to a
local device or directory, the backup options are determined by the local access, as
defined using the Add or Change Device/Directory access to Local Users option.
2-4
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
When you back up to a remote backup server, the backup options are determined
by the Add or Change Client Host Access to this Server option executed on the
server. If no directories are defined for disk image file backups, then no disk file
options are provided.
Figure 2-6 shows the backup options that might be included on the SMIT menu:
Backup Options
Move cursor to desired item and press Enter.
Backup
Backup
Backup
Backup
Backup
the System (Installation Image)
Volume Groups
Filesystems
Logical Volumes
Files or Directories
+--------------------------------------------------------------------------+
|
Select Backup Device or Network Install Image Directory
|
|
|
| Move cursor to desired item and press Enter.
|
|
|
| Dir
/usr/lpp/sysback/images/local
|
| hambone.austin.ibm.com /dev/rmt0
5.0 GB 8mm Tape Drive
|
|
|
| F1=Help
F2=Refresh
F3=Cancel
|
| F8=Image
F10=Exit
Enter=Do
|
F1| /=Find
n=Find Next
|
F9+--------------------------------------------------------------------------+
Figure 2-6. Backup Options
The first column contains the device type and can be any of the following:
VirDev
Virtual devices that you have configured. The device name (vdev0)
is automatically assigned by the system. The description shown is
either one that you assign or a default description assigned by the
system. The physical devices that make up the virtual device are
listed at the end of the line in parentheses.
Tape
Tape drives available on the local system. Local tape drives are
always displayed first in the list.
Dskt
Diskette drives on the local system. A diskette option is not
displayed for a system backup.
Dir
For backup options, a directory name, if the user has permission to
back up to a disk file. The directory shown is the directory or
directories specified when backup permission was enabled for the
user. For list, verify, and restore options, all files in the user’s
backup directories are displayed.
File
For list, verify, recreate and restore options, the name of a local
disk image file. Only files in the directories the user has access to
are displayed. Files of all backup types are shown.
Hostname:
A remote option. The hostname is the name of a server for which
this host has remote access. The information can be any one of the
following:
Chapter 2. SMIT Overview
2-5
Tape Drive Name
Tape drives on the remote machine. Only remote
tape drives on hosts that have granted access to
this host, as well as those that have been defined
by the local host, are included on this selector
screen.
Virtual Device Name
Virtual devices configured on the server. The
virtual device description from the remote machine
and the physical tape drive name(s) are shown.
Only virtual devices for hosts that have granted
access to this host, as well as those that have been
defined by the local host, are included on this
selector screen.
File or Directory
For list, verify, recreate and restore options, a
filename is shown for each disk image backup file
that exists on the server in the directories for which
this host has access. For backup options, a
directory name is shown for all directories this host
has permission to write to.
You must select one of the options provided on the selector screen before
proceeding. After you select an option, you can change the device, filename or
unique ID selected on the dialog screen that follows, but the backup cannot
execute unless you have permission to access the specified device or directory.
Backups to Disk Image Files
To perform a disk file backup, select a directory name from the device selector
screen. A filename is automatically generated as follows:
/Directory/Type.Hostname.UniqueID
The filename contains the following parts:
2-6
Directory
The directory selected.
Type
The type of backup to be performed. One of the following is used:
SB
System Backup
VG
Volume group
LV
Logical Volume
FS
Filesystem
FD
File/Directory
Period (.)
Delimiter.
Hostname
The name of the host performing the backup. This is set to
“localhost” if no hostname is defined.
Period (.)
Delimiter.
UniqueID
A unique ID that, by default, contains the date and time of the
backup in “MMDDhhmm” format. For example, if a backup is
made on Feb. 28 at 3:20 P.M., the unique ID would be “02281520”.
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
You can later change this ID to any other value that might more
accurately describe the backup contents.
Although the filename in SMIT is displayed only as shown above, most backups
actually create more than one file, each named as above, but with an additional
extension. The first file has a “.TOC” extension, indicating the file is the table of
contents of the backup. For a File/Directory backup, the “.TOC” file also contains
the backup data, because there is always only one image in this backup. For all
other backup types, an additional file is created for each logical volume or
filesystem included in the backup, with the logical volume name as the file
extension. This file contains the actual backup data for each logical volume or
filesystem.
Backup, List, Verify and Restore Command Output Screen
The output screen for the backup, list, verify, and restore options is different from
that of other SMIT-executed commands. This screen provides the following
advantages:
v Output is not saved in the SMIT log file (smit.log). This prevents very long file
lists from taking unnecessary space on the disk. You are still provided the option
of writing the output to a log file after the command has completed.
v There is no delay when the command has completed. The delay is normally
required for SMIT processing and writing of the smit.log file.
v Command output (stdout) and error (stderr) are displayed in separate windows.
This provides much cleaner output for commands that provide both standard
output and standard error messages.
v The complete command and the time the command started and ended is
displayed at the top of the screen.
The following is an example of the output of a file/directory backup executed
from SMIT:
COMMAND: mkdirback -hhambone.austin.ibm.com -f/sysback.images/fatrat/backu ...
STARTED: Feb 28 18:37:36 ENDED:
STATUS: Running
Command Output
Generating file list ..
Command Error / Status
NOTE: Estimated megabytes is approximate since data is compressed.
Backing up Files "/tivoli"
Start date is Mon Feb 28 18:36:31 2000
User is root at fatrat
Estimated size is 678 MB
ESTIMATED PROGRESS
-------------+--------------------+------------------+-------------% Complete |
Elapsed Time
|
Megabytes
| Kbytes/Sec
-------------+--------------------+------------------+-------------1
| 0 minutes of 42 |
7 of 678
|
272
Figure 2-7. The output of a file/directory executed from SMIT
Note that the screen is separated into two “windows.” The first is for Command
Output (stdout). The second is for Command Error (stderr) and Status. The
Chapter 2. SMIT Overview
2-7
Command Error section is always used to display error messages but is also often
used for other output, such as file lists, status indicators, and so on.
In the above example, a file/directory backup of the /home/data directory was
performed. The user selected the progress indicator report output type during the
backup, rather than the list of files. The progress indicator shows the estimated
time and amount of data to be backed up, the actual percent, time, and amount
complete, and the average performance in kilobytes per second.
When the command has completed, you can perform the following functions by
pressing the appropriate function key:
F1 Help
Displays a help screen that provides details of the functions you
can perform from this output screen.
F2 Refresh
Erases and redisplays all the information you see. This is necessary
only if there are other commands, such as system console
information, running that output to this terminal and overwrite the
command output.
F3 Return
Returns to the previous SMIT dialog screen.
F4 Select
Toggles between the Command Output and the Command
Error/Status window, enabling each to scroll independently. The
active window is indicated by the arrows (===> window<===).
After you have selected the window to scroll, you can then move
through the output using the following keys:
F5 Save
Page Down
Scroll forward one screen
Page Up
Scroll backward one screen
Down—Arrow/Enter
Scroll backward one line
Up—Arrow
Scroll back one line
Home
Display first screen
End
Display last screen
Save the entire output in a log file. You are asked to enter the
name of the log file. The default file is $HOME/sysback.log, where
$HOME is the user’s home directory. Either enter a new file name
or press Enter to accept the default. The following example shows
a sample output file:
COMMAND:
mkdirback -f/dev/rmt1 -s /home/tj
START TIME: Mar 20 02:22:30
END TIME:
Mar 20 02:25:49
STATUS:
Successful
============== Command Output ==============
Rewinding rmt1 ...
Generating file list ..
File/Directory backup completed successfully.
============== Command Output ==============
Backing up Files */home/tj*
Start date is Thu Mar 20 02:22:33 1997
User is root at sysdev41
Estimated size is 169 MB
2-8
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
ESTIMATED PROGRESS
---------------------------------------------------------% Complete | Elapsed Time | Megabytes | Kbytes/Sec
---------------------------------------------------------100
3 minutes of 3
170 of 170
913
Backup ended Thu Mar 20 02:25:46 1997
170 megabytes written to 1 volume.
SMIT Help Screens
You can obtain help information from any point within SMIT. If a menu selection
is currently displayed, highlight a function and press F1 for information regarding
the use of that function. For any dialog selection, F1 provides information on the
use of that selection, the options available, and how they affect the function being
performed.
SMIT Fastpaths
From any SMIT menu or dialog screen (where you type or change options), press
F8 to display the SMIT fastpath for quickly reaching that selection from the
command line. Later, to skip directly to the desired selection, you can type smit
fastpath at the command line, where fastpath is the fastpath for the desired
selection.
Obtaining Command Line Options from SMIT
You can execute any SMIT command outside of SMIT by placing the command in
a shell script. You can place multiple commands in shell scripts to be executed in
sequence, enabling a sequence of multiple commands to be performed without
user interaction. You can also place the shell scripts in cron, an AIX job scheduling
facility.
To obtain the command line options for a SMIT command, use the following steps:
1. Select the SMIT option to perform and any additional selections required for
this function.
2. Fill out any fields or change the field values in the dialog screen as you would
when executing the command from SMIT.
3. Before pressing Enter to execute the command from SMIT, press F6 to show the
command on the screen.
The following is an example of the SMIT output when you press F6:
Chapter 2. SMIT Overview
2-9
Backup Files or Directories
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
[Entry Fields]
Hostname of server
hambone.austin.ibm.com
Device name
/dev/rmt0
Report output type
progress indicator
+
+--------------------------------------------------------------------------++
|
SHOW COMMAND STRING
|+
|
| #
| Press Enter or Cancel to return to the
|
| application.
|+#
|
|
* | # Ignore "/usr/sbin/sbsmitout -sN" prefix
|+
| # when executing outside of SMIT:
|
|
|
| /usr/sbin/sbsmitout -s7 mkdirback -h’hambone.austin.ibm.com’ -f’/dev/r |
| mt0’ ’-x’ ’/home’
|
|
|
F1| F1=Help
F2=Refresh
F3=Cancel
|
F5| F8=Image
F10=Exit
Enter=Do
|
F9+--------------------------------------------------------------------------+
Figure 2-8. Results of Using the F6 Key to Show the Command String
Single quotes (’) around text are required only when there is more than one word
to a single command argument. They do not need to be used to surround a single
word.
Backup, list, verify and restore commands contain the comment indicating that the
prefix of the command /usr/sbin/sbsmitout -sN should be omitted. This prefix
causes the command to be executed within the two-part SysBack SMIT command
output screen previously described. Do not use this output screen should when
executing commands at the command line if the intention is to not require user
interaction.
2-10
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Chapter 3. License Configuration
In order to license your IBM Tivoli Storage Manager for System Backup and
Recovery (SysBack) software, you must simply install the:
v tivoli.tsm.client.sysback.license.rte
v tivoli.tsm.client.sysback.license.cert
filesets included on your product media in addition to the installation of the base
tivoli.tsm.client.sysback.rte file set.
© Copyright IBM Corp. 2002
3-1
3-2
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Chapter 4. Performing Backups
As soon as SysBack is installed, you can perform a backup to any local device or
to the default SysBack backup image files directories.
To access the Backup Options:
1. At a command line, type smit.
2. Select System Backup & Recovery for AIX.
3. Select Backup & Recovery Options.
You can also access this menu using the fastpath. To do this, type smit sb_backup
at a command line. The following screen is displayed:
Backup Options
Move cursor to desired item and press Enter.
Backup
Backup
Backup
Backup
Backup
F1=Help
F9=Shell
the System (Installation Image)
Volume Groups
Filesystems
Logical Volumes
Files or Directories
F2=Refresh
F10=Exit
F3=Cancel
Enter=Do
F8=Image
Figure 4-1. Backup Options
You can select any of the backup options, each of which are described in detail in
this chapter. The first option, Backup the System (Installation Image), is used to
create a backup of the operating system and any additionally selected volume
groups. This is the only backup from which the system can be reinstalled after an
operating system failure. All other backups options back up smaller increments of
data, such as volume groups or select filesystems.
Understanding Incremental Backups
Incremental backups make it possible to back up only data that has changed since
a prior backup, significantly reducing the backup time and media usage.
Incremental backups can be performed for both volume group and filesystem
backups.
You perform an incremental backup by indicating an incremental backup level.
This level indicates that only files that have changed since the most recent previous
© Copyright IBM Corp. 2002
4-1
level should be included in the backup. Specifying level 0 indicates that all files
should be backed up. Level 0 is still considered an incremental level; it is the base
for all incremental levels that follow.
Any number of incremental levels from 0 through 9 can be used, but using too
many levels can complicate the restore process.
Consider the following incremental backup examples:
Incremental backup of a filesystem is performed as follows:
Monday
level 0
Tuesday
level 1
Wednesday
level 2
Thursday
level 3
Friday
level 4
Saturday
level 5
Sunday
level 6
In this case, if you need to restore the entire filesystem from the backup performed
on Sunday, you must restore every single backup from 0 through 6 in order to
incrementally reapply the changes to date. Always restore level 0 first (when
restoring from scratch) because only level 0 contains a backup of the entire
filesystem. Then, in the correct sequence, reapply each level that follows, adding,
replacing, and removing files until the filesystems return to their most recent state.
Incremental backup of all volume groups, except the rootvg group, is performed
as follows:
1st of month
level 0
every Friday
level 4
Monday-Thursday
level 7
In this example, a volume group backup is performed on the first day of the
month. Every Friday a level 4 is performed, and every other day of the week
(excluding Saturday and Sunday) a level 7 is performed. Because only three levels
are used, no more than three backups need to be restored. If you need to restore
the entire volume group because of failure, the restore sequence would be:
1. Restore level 0.
2. Restore the most recent level 4 if a level 4 was performed since the last level 0.
3. Restore the most recent level 7 if a level 7 was performed since the last level 4
or level 0.
Note: Use care when combining both volume group and filesystem incremental
backups. An incremental volume group backup is identical to performing an
incremental backup of each individual logical volume and filesystem in the
volume group. Therefore, after performing a volume group level 0 backup,
you can perform other incremental backup levels for specific filesystems. If a
level 0 backup of a filesystem is performed after a level 0 of the volume
group containing that filesystem, any higher-level incremental levels for that
filesystem are applied to the most recent (filesystem) backup.
4-2
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
For volume group incremental backups, an entire logical volume is included if any
part of the logical volume has changed. If no data in the logical volume had
changed, the logical volume is not included in the backup.
Understanding Pull Backups
Pull Backups help you centrally manage your backups by enabling you to initiate,
and thereby control, backup operations from a single location. You can further
automate this central management of backups using the SysBack Scheduling and
Scripting functions described in Chapter 7, “Scheduled Backups and Scripts”, on
page 7-1.
The machine that “pulls” the backup is called the “initiator” while the machine
that is backed up is called the “backup system”. The machine that receives that
backup data is called the “destination system”.
Pull backup enablement requires, not only Remote Services configuration, but also
Remote Command Access configuration. The Remote Services configuration
enables the data to be sent across the network from the backup system to the data
destination machine. The Remote Command Access configuration enables the
initiator system to pull the backups from the backup system.
Pull Backups can be done in a two-way or three-way manner. When the initiator
system is also the data destination system, this is a two-way backup. When the
initiator system is a different machine than the data destination system, and
different than the backup system, this is a three-way backup.
Note that for three-way backups, Remote Services must be configured between
Machine A, the backup system, and Machine B, the data destination. Remote
Command Access must be also configured between Machine A, the backup system,
and Machine C, the backup initiator. For more information on configuring Remote
Services and Remote Command Access, see Chapter 9, “Remote Services”, on
page 9-1.
The following SysBack backup options are discussed in this chapter:
v
v
v
v
v
Backing
Backing
Backing
Backing
Backing
up
up
up
up
up
the System (Installation Image)
Volume Groups
Filesystems
Logical Volumes
Files or Directories
Backing up the System (Installation Image)
A system backup can be used in the following ways:
v To reinstall the entire system to its original state.
v To recreate or restore specific volume groups, filesystems, or both on an already
active system.
v To install a system backup created on one machine onto another machine with a
different processor, platform type, or other system devices. This is called cloning.
However, the machine being backed up must have installed all of the device
support for the destination platform type, processor type and other required
devices.
To back up the entire system:
Chapter 4. Performing Backups
4-3
1. From the Backup & Recovery Options menu, select Backup the System
(Installation Image).
Notes:
a. You must be logged in to the system as a root user.
b. At a command line, type smit sb_sysback.
2. If there are volume groups on the system other than rootvg, select the volume
groups to include in the backup. From the list of volume groups, either select
none (rootvg only) or highlight each additional volume group to include and
press F7. Press Enter to continue.
3. On the Device Selector screen, highlight the device you want to use and press
Enter.
4. One of the following screens is displayed. These screens show the default
options and values available when you back up to either a local tape drive or a
disk image file.
Backup the System
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
Hostname of server
Device name
Create a Power Backup?
Report output type
Platform/kernel type for tape boot image
Network install support to include
Include non-JFS logical volumes?
Rewind tape before starting backup?
Forward to End of Tape before Starting Backup
Compress data before writing to media?
User description
Buffer size (in Kbytes)
Preserve physical partition mapping?
Device name for remote volume prompt
Non-rootvg Volume Groups to include
[MORE]
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
[Entry Fields]
hambone.austin.ibm.com
/dev/rmt0
no
+
progress indicator
+
[rspc]
+
[]
+
yes
+
yes
+
no
+
no
+
[]
[64]
+#
no
+
[]
[]
+
F3=Cancel
F7=Edit
Enter=Do
F4=List
F8=Image
Figure 4-2. Default Options Available for Backing Up to Tape
4-4
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Backup the System
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
[Entry Fields]
hambone.austin.ibm.com
/sysback.images/fatrat>
no
[03061826]
no
progress indicator
yes
yes
[]
all hosts
all users
[64]
no
no
[]
Hostname of server
Images directory
Create a Power Backup?
* Backup file ID
Dynamically Created Backup File ID
Report output type
Include non-JFS logical volumes?
Compress data before writing to media?
User description
Host read permission
User read permission
Buffer size (in Kbytes)
Preserve physical partition mapping?
Overwrite existing backup with same ID?
Non-rootvg Volume Groups to include
[MORE]
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
F3=Cancel
F7=Edit
Enter=Do
+
+
+
+
+
+
+
+#
+
+
+
F4=List
F8=Image
Figure 4-3. Default Options Available for Backing Up to Disk Image File
The fields have the following definitions:
Hostname of server
The hostname of the data destination system on which the
selected backup device or file exists. You cannot change this
field. To select a different hostname, press F3 to return to the
menu and reselect this option to obtain a list of hosts and
device options.
Device name
The device, if a tape drive or virtual device was selected. You
cannot change this field from this screen.
Images directory
The directory name, if previously selected. You cannot change
this field from this screen.
Create a Power Backup?
Default=no. Use Tab to change this field to “yes” if you want
this backup to be a “Power” backup. A Power backup backs up
all filesystems as “raw” logical volumes, which usually
provides much better backup and restore performance, but
imposes the following limitations:
v You can restore only raw logical volumes from this backup.
Therefore, when choosing to restore data from the backup,
you will be allowed to restore only an entire logical volume
and only if it is inactive. Because of this, you cannot restore
any individual files or directories, only entire filesystems;
and no filesystems, such as /, /usr and /var can be restored
since they are always active.
v Because all filesystems are backed up as raw logical volumes,
the entire logical volume is backed up even if only
one–fourth of the filesystem is used. If this were the case
with all filesystems, the backup and installation would likely
take longer than normal.
Chapter 4. Performing Backups
4-5
v When installing or recreating volume groups or logical
volumes from this backup, you cannot change the following
attributes:
–
–
–
–
Any filesystem attributes
Logical volume name
Logical volume size
The volume group in which you want to create a logical
volume
Backup file ID
Default=date/time (MMDDhhmm). Uniquely identifies the
backup. You can change this field to any value but it cannot
contain single (’) or double (″) quotes. This field is active only
when you back up to a disk image file.
Dynamically Created Backup File ID
This option lets you generate a dynamic backup file ID. It uses
the same current date and time values at the time of command
execution as the ID. If the value of the Backup File ID field set
to yes, it will overwrite the Backup File ID field.
Report output type
Default=progress indicator. Keep the default value if you want
to display a progress indicator during the backup process,
which indicates the approximate total backup time and amount
completed. Press Tab to select “file list” if you want to display
a list of files as they are being backed up, or Tab once more to
show “only errors” that occur during the backup.
Platform/kernel type for tape boot image
Default=current platform/kernel type. This field is displayed
only when you use a tape drive or virtual device. The field
does not display a value if the current platform does not
support tape boot or tape boot support is not installed. Change
this field only if you are creating a backup that you want to
install (clone) onto another platform type or to a machine
running a different kernel. Press F4 to list the platforms/kernels
that have tape boot support installed on the system and select
from the list.
The following platform/kernel types are currently supported:
chrp
Common Hardware Reference Platform
chrp/MP
Multiprocessor Common Hardware Reference
Platform
rs6k
RISC System/6000 (uniprocessor)
rs6k/MP
Multiprocessor RISC System/6000
rspc
PCI-based (PC) RISC System/6000
(uniprocessor)
rspc/MP
Multiprocessor PCI-based (PC) RISC
System/6000
The first part of the selection is the platform type. The second
part (/MP) indicates the kernel type. If the kernel type portion
is omitted, a uniprocessor (single processor) kernel is assumed.
4-6
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Network install support to include
Default=(Blank/None). If, after starting a system from this
backup tape, you want the ability to perform a network
installation (rather than installing from this backup), press F4 to
display a list of supported network types and select one from
the list. You must select the network type of the adapter that
you want to use to install the system from the network server.
Include non-JFS logical volumes?
Default=yes. To exclude all non-journaled filesystems (raw
logical volumes) from the backup, press Tab to change the
value to “no”. If you select “no” to this option, all raw logical
volume data is excluded from the backup, but the logical
volumes can be recreated using the LVM information on the
backup. If you select “yes”, all raw logical volumes are
included on the backup unless you specifically excluded them
in the exclude list.
Rewind tape before starting backup?
Default=yes. Press Enter to rewind the tape before the backup
begins. If you are placing multiple backup images onto a single
tape, then select “no” to place this backup on the tape at the
current location. This option is displayed only when you back
up to a tape device. For a system backup, tape boot information
is placed on the tape at the beginning of the first backup. This
information is not placed on the tape for subsequent backups.
Note: If this is the first backup to this tape, or if you have
restarted the system, reinserted the tape, or changed the
tape drive block size, you should select to rewind the
tape. Only if you completed another backup and have
not performed any of these functions will the tape be
positioned correctly at the end of the previous backup
image.
Forward to End of Tape Before Starting Backup
This option lets you fast forward to the end of the last SysBack
backup on the tape for all backup types. Use this option when
stacking multiple backups on the same tape that has been
rewound or ejected. This options should also be used when the
system has been restarted or the block size of the tape driver
has been changed from the last backup. This eliminates having
to manually position the tape before stacking a backup image.
Compress data before writing to media?
Default=“no” for tapes and “yes” for disk image backups. Press
Tab to change the option. “yes” indicates that the data will be
compressed before it is written to the backup media.
Compressing the data usually increases backup performance
and uses between 25 and 40% less space on the media. For best
performance, do not select this option if the backup device
provides its own hardware data compression. Compressing
data with the software uses considerably more CPU cycles to
process the data.
User description
Default=(Blank/None). Enter up to 60 characters of descriptive
Chapter 4. Performing Backups
4-7
information to add to the volume label on the backup. Do not
use single (’) or double (″) quotation marks in the description.
Host read permission
Default=all hosts. By default, all hosts with access to the system
on which this file is written have permission to read, or restore
from this image. Press Tab to limit read permission to the
original host or the server on which the file is written. This
option is active only when you back up to a disk image file.
User read permission
Default=all users. By default, any user on a host with access to
the system and directory on which this file is written has
permission to read, or restore from, this image. Press Tab to
limit the read permission to only the original user, the root user
on the original host, or the server to which the file is written.
This option is active only when you back up to a disk image
file.
Buffer size (In Kbytes)
Default=64. Changing this value either increases or decreases
the amount of data that is written to the output device in a
single I/O operation. Some devices with small buffers might
require you to reduce this value, while other devices with very
large buffers can benefit from increased performance by
increasing this value. If you are unsure, use the default value.
This value must be a multiple of the tape block size.
Preserve physical partition mapping?
Default=no. Selecting “yes” to this option ensures that all
logical volumes will be recreated from this backup using the
same physical partitions that are currently allocated to them.
Unless the logical volumes were originally created using
physical partition maps, it is generally recommended that they
not be recreated using the same partitions, because this
preserves fragmentation that develops over time as logical
volumes and filesystems expand.
When you recreate a volume group or logical volume or
perform a system installation, you can choose whether or not to
use the prior partition map. If the original physical volumes do
not exist or are not large enough to contain the same partitions,
the logical volumes will be recreated without using the maps.
Device name for remote volume prompt
Default=(Blank/None). Enter a TTY, LFT or PTS device name to
send the volume prompt to a specified device rather than to the
current SMIT screen. If you are backing up to a remote host,
the device name you specify is attached to the remote host.
This field is displayed only when the backup device is a tape
drive or virtual device.
Examples of device names are /dev/tty0, /dev/lft0 and /dev/pts/5.
You can determine the terminal device name by typing tty at
the command line on that device.
Overwrite existing backup with same ID?
Default=no. If the Backup File ID field contains a value already
used for an existing system backup, change this field to “yes”
by pressing Tab. Doing so indicates that you want to overwrite
4-8
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
the previous backup with this backup. After you begin this
backup, the prior disk image files are removed and cannot be
recovered. This field is displayed only when you back up to a
disk image file.
Non-rootvg Volume Groups to include
Default displays the volume groups previously selected, if any.
You can add or delete volume groups from this list to include
or exclude them from the backup. You can also press F4 to list
the available volume groups and select from the list using F7.
Hostname to Backup
The hostname of the backup system, the machine that will send
its data to the data destination server.
5. After verifying all selections made above, press Enter to begin the backup.
If you selected to back up to tape, the tape block size is automatically changed to
512-byte blocks for writing the boot and installation images on the tape. This is
necessary for the installation process to properly read the tape. After the boot and
installation images are written to the tape, the block size is changed back to the
original value and the remainder of the backup is written.
This process calculates and records certain information about your system
configuration for later use by the installation process. The boot images are then
placed on the tape, followed by each filesystem or logical volume in the selected
volume groups.
Pre-Backup Script: If you created a custom script to run before the system backup,
it runs first. This script might mail a message to users, remove temporary files,
stop applications, or other similar functions. Refer to Appendix C, “Creating Scripts
for Customizing the System Backup and Install Process”, on page C-1 for
information on creating pre-backup scripts.
Post-Backup Script: If you created a custom script to run after the completion of
the system backup, it runs after the backup is complete and before the tape is
rewound and verified. This script might mail a message to users or re-enable
applications that were disabled prior to the backup starting. Refer to Appendix C,
“Creating Scripts for Customizing the System Backup and Install Process”, on
page C-1 for information on creating post-backup scripts.
Backing up Volume Groups
If your system is organized so that the physical volumes, or hard disks, are placed
into separate volume groups, then you probably want to back up these volume
groups independently, for the following reasons.
First, data in different volume groups might require backups to be performed at
different frequencies. For instance, critical data changed daily should be backed up
at least once daily. Application programs or history files might not change on a
daily basis and might be only backed up on a weekly or monthly basis.
Secondly, the most common hardware failure in any computer system is a hard
disk failure. If a single disk fails within a volume group, you want to replace the
disk and restore only the data in that one volume group. Separately backing up
each volume group enables you to restore only a single volume group without
touching the rest of the data on the system.
Chapter 4. Performing Backups
4-9
You can use the Backup Volume Groups option to back up one or more volume
groups. All filesystems and logical volumes in the selected volume groups are
included on the backup unless they are specified in the exclude list.
Note: Do not use this option to back up the files in the rootvg volume group. To
create a backup of the rootvg volume group or all volume groups, which
can be used to reinstall the system, use the Backup the System (Installation
Image) option. If you use the Backup Volume Groups option to backup the
rootvg volume group, you can restore individual files from the backup but
cannot reinstall the system.
Files, directories, or logical volumes contained in the exclude list are not included in
the backup. If a filesystem mount point is contained in the exclude list, the entire
filesystem is ignored. However, the information pertaining to all filesystems and
logical volumes is recorded to enable the recreation of the filesystems or logical
volumes as requested.
Before performing a volume group backup, log in to the system as a root user. This
ensures that you have the appropriate permissions to read all of the files in the
volume groups.
To back up one or more volume groups:
1. From the Backup Options menu, select Backup Volume Groups.
Note: From a command line, type smit sb_mkvgback.
2. Select the volume group or groups to include in the backup from the list
displayed. To select a single group, highlight the group and press Enter. To
select multiple groups, highlight each group and press F7. Then, press Enter to
continue.
3. On the Device Selector screen, highlight the device you want to use and press
Enter.
4. One of the following screens is displayed. These screens show the default
options and values available when you back up to either a local tape drive or a
disk image file.
4-10
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Backup a Volume Group
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
Hostname of server
Device name
Report output type
Rewind tape before starting backup?
Forward to End of Tape Before Starting Backup
Compress data before writing to media?
Incremental backup level (0=full)
Include non-JFS logical volumes?
User description
Buffer size (in Kbytes)
Preserve physical partition mapping?
Device name for remote volume prompt
* Volume Group name(s)
[MORE]
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
F3=Cancel
F7=Edit
Enter=Do
[Entry Fields]
hambone.austin.ibm.com
/dev/rmt0
progress indicator
+
yes
+
no
+
no
+
[]
#
yes
+
[]
[64]
+#
no
+
[]
[datavg]
+
F4=List
F8=Image
Figure 4-4. Default Options and Values Available When Backing Up to a Local Tape Drive
Backup a Volume Group
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
Hostname of server
Images directory
* Backup file ID
Dynamically Created Backup File ID
Report output type
Compress data before writing to media?
Incremental backup level (0=full)
Include non-JFS logical volumes?
User description
Host read permission
User read permission
Buffer size (in Kbytes)
Preserve physical partition mapping?
Overwrite existing backup with same ID?
* Volume Group name(s)
[MORE]
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
F3=Cancel
F7=Edit
Enter=Do
[Entry Fields]
hambone.austin.ibm.com
/sysback.images/fatrat>
[03071030]
no
progress indicator
yes
[]
yes
[]
all hosts
all users
[64]
no
no
[datavg]
+
+
+
#
+
+
+
+#
+
+
+
F4=List
F8=Image
Figure 4-5. Default Options and Values Available When Backing Up to Disk Image File
The fields have the following definitions:
Hostname of server:
The hostname of the data destination system on which the
selected backup device or file exists. You cannot change this
field. To select a different hostname, press F3 to return to the
menu and reselect this option to obtain a list of hosts and
device options.
Chapter 4. Performing Backups
4-11
Device name
The device, if a tape drive or virtual device was selected. You
cannot change this field from this screen.
Images directory
The directory name, if previously selected. You cannot change
this field from this screen.
Backup file ID
Default=date/time (MMDDhhmm). Uniquely identifies the
backup. You can change this field to any value but it cannot
contain single (’) or double (″) quotes. This field is active only
when you back up to a disk image file.
Dynamically Created Backup File ID
Generates a dynamic backup file ID. This option uses the same
current date and time values at the time of command execution
as the ID. If a dynamic backup file ID already exists and this
option is set to “yes”, the existing ID is overwritten.
Report output type
Default=progress indicator. Keep the default value if you want
to display a progress indicator during the backup process,
which indicates the approximate total backup time and amount
completed. Press Tab to select “file list” if you want to display
a list of files as they are being backed up, or Tab once more to
show “only errors” that occur during the backup.
Compress data before writing to media?
Default=“no” for tapes and “yes” for disk image backups. Press
Tab to change the option. “yes” indicates that the data will be
compressed before it is written to the backup media.
Compressing the data usually increases backup performance
and uses between 25 and 40% less space on the media. For best
performance, do not select this option if the backup device
provides its own hardware data compression. Compressing
with software data uses considerably more CPU cycles to
process the data.
Incremental backup level (0=full)
Default=(Blank/None). To perform an incremental backup of
the volume group, enter an incremental level. Level 0 indicates
a backup of all files and logical volumes. Leaving this field
blank also backs up all files and logical volumes but the
previous level 0 incremental backup information, if any, is not
overwritten. Refer to “Understanding Incremental Backups” on
page 4-1 for additional details.
Include non-JFS logical volumes?
Default=yes. To exclude all non-journaled filesystems (raw
logical volumes) from the backup, press Tab to change the
value to “no”. If you select “no” to this option, all raw logical
volume data is excluded from the backup, but the logical
volumes can be recreated using the LVM information on the
backup. If you select “yes”, all raw logical volumes are
included on the backup unless you specifically excluded them
in the exclude list.
User description
Default=(Blank/None). Enter up to 60 characters of descriptive
4-12
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
information to add to the volume label on the backup. Do not
use single (’) or double (″) quotation marks in the description.
Host read permission
Default=all hosts. By default, all hosts with access to the system
on which this file is written have permission to read, or restore
from, this image. Press Tab to limit read permission to the
original host or the server on which the file is written. This
option is displayed only when you back up to a disk image file.
User read permission
Default=all users. By default, any user on a host with access to
the system and directory on which this file is written has
permission to read, or restore from, this image. Press Tab to
limit the read permission to only the original user, the root user
on the original host, or the server to which the file is written.
This option is displayed only when you back up to a disk
image file.
Rewind tape before starting backup?
Default=yes. Press Enter to rewind the tape before the backup
begins. If you are placing multiple backup images onto a single
tape, then select “no” to place this backup on the tape at the
current location. This option is displayed only when you back
up to a tape device. For a system backup, tape boot information
is placed on the tape at the beginning of the first backup. This
information is not placed on the tape for subsequent backups.
Note: If this is the first backup to this tape, or if you have
restarted the system, reinserted the tape, or changed the
tape drive block size, you should select to rewind the
tape. Only if you completed another backup and have
not performed any of these functions will the tape be
positioned correctly at the end of the previous backup
image.
Forward to End of Tape Before Starting Backup
Forwards to the end of the last SysBack backup on the tape for
all backup types. Use this option when stacking multiple
backups on a tape that has been rewound or ejected. Also use
this option options when the system has been restarted or
when the block size of the tape drive has been changed from
the last backup.
Buffer size (In Kbytes)
Default=64. Changing this value either increases or decreases
the amount of data that is written to the output device in a
single I/O operation. Some devices with small buffers might
require you to reduce this value, while other devices with very
large buffers can benefit from increased performance by
increasing this value. If you are unsure, use the default value.
This value must be a multiple of the tape block size.
Preserve physical partition mapping?
Default=no. Selecting “yes” to this option ensures that all
logical volumes will be recreated from this backup using the
same physical partitions that are currently allocated to them.
Unless the logical volumes were originally created using
physical partition maps, it is generally recommended that they
Chapter 4. Performing Backups
4-13
not be recreated using the same partitions, because this
preserves fragmentation that develops over time as logical
volumes and filesystems expand.
When you recreate a volume group or logical volume or
perform a system installation, you can choose whether or not to
use the prior partition map. If the original physical volumes do
not exist or are not large enough to contain the same partitions,
the logical volumes will be recreated without using the maps.
Device name for remote volume prompt
Default=(Blank/None). Enter a TTY, LFT or PTS device name to
send the volume prompt to a specified device rather than to the
current SMIT screen. If you are backing up to a remote host,
the device name you specify is attached to the remote host.
This field is displayed only when the backup device is a tape
drive or virtual device.
Examples of device names are /dev/tty0, /dev/lft0 and /dev/pts/5.
You can determine the terminal device name by typing tty at
the command line on that device.
Overwrite existing backup with same ID?
Default=no. If the Backup File ID field contains a value already
used for an existing system backup, change this field to “yes”
by pressing Tab. Doing so indicates that you want to overwrite
the previous backup with this backup. After you begin this
backup, the prior disk image files are removed and cannot be
recovered. This field is displayed only when you back up to a
disk image file.
Volume Group name(s)
The volume groups selected on the first selection screen, each
separated by spaces. You can change the list by adding or
removing volume group names, or you can press F4 to list the
volume groups on the system and F7 to select each volume
group from the list. All volume groups indicated in this field
are included in this backup.
Hostname to Backup
The hostname of the backup system, the machine that will send
its data to the data destination server.
5. After confirming the above entries, make sure the backup media is properly
inserted in the drive and press Enter to begin the backup.
Note: If you are backing up to a tape drive, the tape will NOT be rewound upon
completion of the backup. This enables you to perform additional backups
(also without rewinding), thereby “stacking” the backups onto a single tape.
Backing up Filesystems
A filesystem backup can only be performed for mounted (active) filesystems.
Various filesystems are often backed up at different intervals from each other or
from the volume group or system backup, because the data in different filesystems
is changed at different intervals.
Because filesystems are contained in the system and volume group backups, you
do not need to back them up separately, unless you want to back up this data
independently or more frequently.
4-14
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
All files and directories in the filesystems are included in the backup unless you
specify them in the exclude list. If the mount point of a filesystem to be backed up
is also in the exclude list, the backup of the filesystem continues, but no files in the
filesystem are included.
Before performing a filesystem backup, log in to the system as a root user. This
ensures that you have the appropriate permissions to read all of the files in the
filesystems.
To back up one or more filesystems:
1. From the Backup Options menu, select Backup Filesystems.
Note: From a command line, type smit sb_mkfsback.
2. Select a filesystem from the list of filesystem mount points. To select a single
filesystem, highlight a directory and press Enter. To select multiple filesystems,
highlight each directory and press F7. When you’ve made all your selections,
press Enter.
3. On the Device Selector screen, highlight the device you want to use and press
Enter.
4. One of the following screens is displayed. These screens show the default
options and values available when you back up to either a local tape drive or a
disk image file.
Backup a Filesystem
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
Hostname of server
Device name
Report output type
Forward to End of Tape Before Starting Backup
Rewind tape before starting backup?
Compress data before writing to media?
Incremental backup level (0=full)
User description
Buffer size (in Kbytes)
Preserve physical partition mapping?
Device name for remote volume prompt
* Filesystem(s) (directory names)
[MORE]
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
F3=Cancel
F7=Edit
Enter=Do
[Entry Fields]
hambone.austin.ibm.com
/dev/rmt0
progress indicator
+
no
+
yes
+
no
+
[]
#
[]
[64]
+#
no
+
[]
[/tivoli]
+/
F4=List
F8=Image
Figure 4-6. Default Options and Values Available When Backing Up to a Local Tape Drive
Chapter 4. Performing Backups
4-15
Backup a Filesystem
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
[Entry Fields]
hambone.austin.ibm.com
/sysback.images/fatrat>
[03071032]
no
progress indicator
yes
[]
[]
all hosts
all users
[64]
no
no
[/tivoli]
Hostname of server
Images directory
* Backup file ID
Dynamically Created Backup File ID
Report output type
Compress data before writing to media?
Incremental backup level (0=full)
User description
Host read permission
User read permission
Buffer size (in Kbytes)
Preserve physical partition mapping?
Overwrite existing backup with same ID?
* Filesystem(s) (directory names)
[MORE]
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
F3=Cancel
F7=Edit
Enter=Do
+
+
+
#
+
+
+#
+
+
+/
F4=List
F8=Image
Figure 4-7. Default Options and Values Available When Backing Up to Disk Image File
The fields have the following definitions:
Hostname of server
This field contains the hostname of the backup server on which
the selected backup device or file exists. You may not change
this field. To select a different hostname, press F3 to return to
the menu and reselect this option to obtain a list of hosts and
device options.
Device name
The device, if a tape drive or virtual device was selected. You
cannot change this field from this screen.
Images directory
The directory name, if previously selected. You cannot change
this field from this screen.
Create a Power Backup?
Default=no. Use Tab to change this field to “yes” if you want
this backup to be a “Power” backup. A Power backup backs up
all filesystems as “raw” logical volumes, which usually
provides much better backup and restore performance, but
imposes the following limitations:
v You can restore only raw logical volumes from this backup.
Therefore, when choosing to restore data from the backup,
you will be allowed to restore only an entire logical volume
and only if it is inactive. Because of this, you cannot restore
any individual files or directories, only entire filesystems;
and no filesystems, such as /, /usr and /var can be restored
since they are always active.
v Because all filesystems are backed up as raw logical volumes,
the entire logical volume is backed up even if only
one–fourth of the filesystem is used. If this were the case
with all filesystems, the backup and installation would likely
take longer than normal.
4-16
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
v When installing or recreating volume groups or logical
volumes from this backup, you cannot change the following
attributes:
–
–
–
–
Any filesystem attributes
Logical volume name
Logical volume size
The volume group in which you want to create a logical
volume
Backup file ID
Default=date/time (MMDDhhmm). Uniquely identifies the
backup. You can change this field to any value but it cannot
contain single (’) or double (″) quotes. This field is active only
when you back up to a disk image file.
Dynamically Created Backup File ID
Generates a dynamic backup file ID. This option uses the same
current date and time values at the time of command execution
as the ID. If a dynamic backup file ID already exists and this
option is set to “yes”, the existing ID is overwritten.
Report output type
Default=progress indicator. Keep the default value if you want
to display a progress indicator during the backup process,
which indicates the approximate total backup time and amount
completed. Press Tab to select “file list” if you want to display
a list of files as they are being backed up, or Tab once more to
show “only errors” that occur during the backup.
Compress data before writing to media?
Default=“no” for tapes and “yes” for disk image backups. Press
Tab to change the option. “yes” indicates that the data will be
compressed before it is written to the backup media.
Compressing the data usually increases backup performance
and uses between 25 and 40% less space on the media. For best
performance, do not select this option if the backup device
provides its own hardware data compression. Compressing
data with software uses considerably more CPU cycles to
process the data.
Incremental backup level (0=full)
Default=(Blank/None). To perform an incremental backup of
the volume group, enter an incremental level. Level 0 indicates
a backup of all files and logical volumes. Leaving this field
blank also backs up all files and logical volumes but the
previous level 0 incremental backup information, if any, is not
overwritten. Refer to “Understanding Incremental Backups” on
page 4-1 for additional details.
User description
Default=(Blank/None). Enter up to 60 characters of descriptive
information to add to the volume label on the backup. Do not
use single (’) or double (″) quotation marks in the description.
Host read permission
Default=all hosts. By default, all hosts with access to the system
on which this file is written have permission to read, or restore
from, this image. Press Tab to limit read permission to the
Chapter 4. Performing Backups
4-17
original host or the server on which the file is written. This
option is displayed only when you back up to a disk image file.
User read permission
Default=all users. By default, any user on a host with access to
the system and directory on which this file is written has
permission to read, or restore from, this image. Press Tab to
limit the read permission to only the original user, the root user
on the original host, or the server to which the file is written.
This option is displayed only when you back up to a disk
image file.
Rewind tape before starting backup?
Default=yes. Press Enter to rewind the tape before the backup
begins. If you are placing multiple backup images onto a single
tape, then select “no” to place this backup on the tape at the
current location. This option is displayed only when you back
up to a tape device. For a system backup, tape boot information
is placed on the tape at the beginning of the first backup. This
information is not placed on the tape for subsequent backups.
Note: If this is the first backup to this tape, or if you have
restarted the system, reinserted the tape, or changed the
tape drive block size, you should select to rewind the
tape. Only if you completed another backup and have
not performed any of these functions will the tape be
positioned correctly at the end of the previous backup
image.
Forward to End of Tape Before Starting Backup
Forwards to the end of the last SysBack backup on the tape for
all backup types. Use this option when stacking multiple
backups on a tape that has been rewound or ejected. Also use
this option when the system has been restarted or when the
block size of the tape driver has been changed from the last
backup.
Buffer size (In Kbytes)
Default=64. Changing this value either increases or decreases
the amount of data that is written to the output device in a
single I/O operation. Some devices with small buffers might
require you to reduce this value, while other devices with very
large buffers can benefit from increased performance by
increasing this value. If you are unsure, use the default value.
This value must be a multiple of the tape block size.
Preserve physical partition mapping?
Default=no. Selecting “yes” to this option ensures that all
logical volumes will be recreated from this backup using the
same physical partitions that are currently allocated to them.
Unless the logical volumes were originally created using
physical partition maps, it is generally recommended that they
not be recreated using the same partitions, because this
preserves fragmentation that develops over time as logical
volumes and filesystems expand.
When you recreate a volume group or logical volume or
perform a system installation, you can choose whether or not to
use the prior partition map. If the original physical volumes do
4-18
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
not exist or are not large enough to contain the same partitions,
the logical volumes will be recreated without using the maps.
Device name for remote volume prompt
Default=(Blank/None). Enter a TTY, LFT or PTS device name to
send the volume prompt to a specified device rather than to the
current SMIT screen. If you are backing up to a remote host,
the device name you specify is attached to the remote host.
This field is displayed only when the backup device is a tape
drive or virtual device.
Examples of device names are /dev/tty0, /dev/lft0 and /dev/pts/5.
You can determine the terminal device name by typing tty at
the command line on that device.
Overwrite existing backup with same ID?
Default=no. If the Backup File ID field contains a value already
used for an existing system backup, change this field to “yes”
by pressing Tab. Doing so indicates that you want to overwrite
the previous backup with this backup. After you begin this
backup, the prior disk image files are removed and cannot be
recovered. This field is displayed only when you back up to a
disk image file.
Filesystem(s) (directory names)
The filesystem mount points (directories) selected on the first
selection screen, each separated by spaces. You can change the
list by adding or removing filesystems, or you can press F4 to
list the mounted filesystems on the system and use F7 to select
one or more from the list. All filesystems in this field, unless
included in the exclude list, are included in this backup.
Directory names should be specified relative to the root (/)
directory.
Hostname to Backup
The hostname of the backup system, the machine that will send
its data to the data destination server.
5. After confirming the above entries, make sure the backup media is properly
inserted in the drive and press Enter to begin the backup.
Note: If you are backing up to a tape drive, the tape will NOT be rewound upon
completion of the backup. This enables additional backups to be performed
(also without rewinding), thereby “stacking” the backups onto a single tape.
Backing up Logical Volumes
This option is typically used to back up logical volumes that contain
non-filesystem data, such as data written by various database products or custom
applications.
Logical volumes are contained in the system and volume group backups, so you
do not need to back them up separately, unless you want to back up this data
independently or more frequently.
You can use this option to backup a filesystem logical volume only if the filesystem
is unmounted before the backup is performed. For non-filesystem logical volumes,
the backup can be performed even if the logical volume is active, but it is
Chapter 4. Performing Backups
4-19
recommended that data not be changed in the logical volume during the backup to
ensure the consistency of relational data during the backup.
Since logical volume names must be explicitly selected using this option, they are
backed up even if they are in the exclude list.
To backup one or more logical volumes:
1. From the Backup Options menu, select Backup Logical Volumes.
Note: From a command line, type smit sb_mklvback.
2. Select a logical volume from the list of non-filesystem logical volumes and
logical volumes with unmounted filesystem. To select a single logical volume,
highlight the name and press Enter. To select multiple logical volumes,
highlight each volume group and press F7. When you’ve made all your
selections, press Enter.
3. On the Device Selector screen, highlight the device you want to use and press
Enter.
4. One of the following screens is displayed. These screens show the default
options and values available when you back up to either a local tape drive or a
disk image file.
Backup a Logical Volume
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
Hostname of server
Device name
Show progress indicator?
Rewind tape before starting backup?
Forward to End of Tape Before Starting Backup
Compress data before writing to media?
User description
Buffer size (in Kbytes)
Preserve physical partition mapping?
Device name for remote volume prompt
* Logical Volume name(s)
[MORE]
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
F3=Cancel
F7=Edit
Enter=Do
[Entry Fields]
hambone.austin.ibm.com
/dev/rmt0
yes
+
yes
+
no
+
no
+
[]
[64]
+#
no
+
[]
[lv00]
+
F4=List
F8=Image
Figure 4-8. Default Options and Values Available When Backing Up to Tape
4-20
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Backup a Logical Volume
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
[Entry Fields]
hambone.austin.ibm.com
/sysback.images/fatrat>
[03071034]
no
yes
yes
[]
all hosts
all users
[64]
no
no
[lv00]
Hostname of server
Images directory
* Backup file ID
Dynamically Created Backup File ID
Show progress indicator?
Compress data before writing to media?
User description
Host read permission
User read permission
Buffer size (in Kbytes)
Preserve physical partition mapping?
Overwrite existing backup with same ID?
* Logical Volume name(s)
[MORE]
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
F3=Cancel
F7=Edit
Enter=Do
+
+
+
+
+
+#
+
+
+
F4=List
F8=Image
Figure 4-9. Default Options and Values Available When Backing Up to Disk Image File
The fields have the following definitions:
Hostname of server:
The hostname of the data destination system on which the
selected backup device or file exists. You cannot change this
field. To select a different hostname, press F3 to return to the
menu and reselect this option to obtain a list of hosts and
device options.
Device name
The device, if a tape drive or virtual device was selected. You
cannot change this field from this screen.
Images directory
The directory name, if previously selected. You cannot change
this field from this screen.
Backup file ID
Default=date/time (MMDDhhmm). Uniquely identifies the
backup. You can change this field to any value but it cannot
contain single (’) or double (″) quotes. This field is active only
when you back up to a disk image file.
Dynamically Created Backup File ID
Generates a dynamic backup file ID. This option uses the same
current date and time values at the time of command execution
as the ID. If a dynamic backup file ID already exists and this
option is set to “yes”, the existing ID is overwritten.
Show progress indicator?
Default=“yes”. Press Tab to change the value to “no” if you do
not want the progress indicator shown on the screen during the
backup.
Compress data before writing to media?
Default=“no” for tapes and “yes” for disk image backups. Press
Tab to change the option. “yes” indicates that the data will be
compressed before it is written to the backup media.
Chapter 4. Performing Backups
4-21
Compressing the data usually increases backup performance
and uses between 25 and 40% less space on the media. For best
performance, do not select this option if the backup device
provides its own hardware data compression. Compressing
data with software uses considerably more CPU cycles to
process the data.
User description
Default=(Blank/None). Enter up to 60 characters of descriptive
information to add to the volume label on the backup. Do not
use single (’) or double (″) quotation marks in the description.
Host read permission
Default=all hosts. By default, all hosts with access to the system
on which this file is written have permission to read, or restore
from, this image. Press Tab to limit read permission to the
original host or the server on which the file is written. This
option is displayed only when you back up to a disk image file.
User read permission
Default=all users. By default, any user on a host with access to
the system and directory on which this file is written has
permission to read, or restore from, this image. Press Tab to
limit the read permission to only the original user, the root user
on the original host, or the server to which the file is written.
This option is displayed only when you back up to a disk
image file.
Rewind tape before starting backup?
Default=yes. Press Enter to rewind the tape before the backup
begins. If you are placing multiple backup images onto a single
tape, then select “no” to place this backup on the tape at the
current location. This option is displayed only when you back
up to a tape device. For a system backup, tape boot information
is placed on the tape at the beginning of the first backup. This
information is not placed on the tape for subsequent backups.
Note: If this is the first backup to this tape, or if you have
restarted the system, reinserted the tape, or changed the
tape drive block size, you should select to rewind the
tape. Only if you completed another backup and have
not performed any of these functions will the tape be
positioned correctly at the end of the previous backup
image.
Forward to End of Tape Before Starting Backup
Forwards to the end of the last SysBack backup on the tape for
all backup types. Use this option when stacking multiple
backups on a tape that has been rewound or ejected. Also use
this option when the system has been restarted or when the
block size of the tape driver has been changed from the last
backup.
Buffer size (In Kbytes)
Default=64. Changing this value either increases or decreases
the amount of data that is written to the output device in a
single I/O operation. Some devices with small buffers might
require you to reduce this value, while other devices with very
large buffers can benefit from increased performance by
4-22
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
increasing this value. If you are unsure, use the default value.
This value must be a multiple of the tape block size.
Preserve physical partition mapping?
Default=no. Selecting “yes” to this option ensures that all
logical volumes will be recreated from this backup using the
same physical partitions that are currently allocated to them.
Unless the logical volumes were originally created using
physical partition maps, it is generally recommended that they
not be recreated using the same partitions, because this
preserves fragmentation that develops over time as logical
volumes and filesystems expand.
When you recreate a volume group or logical volume or
perform a system installation, you can choose whether or not to
use the prior partition map. If the original physical volumes do
not exist or are not large enough to contain the same partitions,
the logical volumes will be recreated without using the maps.
Device name for remote volume prompt
Default=(Blank/None). Enter a TTY, LFT or PTS device name to
send the volume prompt to a specified device rather than to the
current SMIT screen. If you are backing up to a remote host,
the device name you specify is attached to the remote host.
This field is displayed only when the backup device is a tape
drive or virtual device.
Examples of device names are /dev/tty0, /dev/lft0 and /dev/pts/5.
You can determine the terminal device name by typing tty at
the command line on that device.
Overwrite existing backup with same ID?
Default=no. If the Backup File ID field contains a value already
used for an existing system backup, change this field to “yes”
by pressing Tab. Doing so indicates that you want to overwrite
the previous backup with this backup. After you begin this
backup, the prior disk image files are removed and cannot be
recovered. This field is displayed only when you back up to a
disk image file.
Logical Volume name(s)
The logical volume name or names selected on the first
selection screen, each separated by spaces. You can change the
list by adding or removing logical volume names, or press F4
to list the valid logical volumes on the system and use F7 to
select one or more from the list.
Hostname to Backup
The hostname of the backup system, the machine that will send
its data to the data destination server.
5. After confirming the above entries, make sure the backup media is properly
inserted in the drive and press Enter to begin the backup.
Note: If you are backing up to a tape drive, the tape will NOT be rewound upon
completion of the backup. This enables additional backups to be performed
(also without rewinding), thereby “stacking” the backups onto a single tape.
During the backup, the Progress Indicator will be displayed, indicating the
estimated backup time, and the amount completed.
Chapter 4. Performing Backups
4-23
Backing up Files or Directories
Use this option to back up a single file or directory or a group of files or
directories. You can also back up only the files that have changed within a certain
number of days. This enables the files or directories specified to be backed up
incrementally from day to day. For instance, you might choose to back up all files
in a directory each Friday evening. Then, for each following day, you can choose to
back up only the files that have changed in the last day. Keep in mind the
following when using this option:
v One day equals exactly 24 hours. Therefore, if you perform a backup of a file at
1:00 am on Tuesday, and the same file is not read until 3:00 am on Wednesday, it
will not be backed up if it was changed at 2:00 am on Tuesday, 25 hours ago. If
files are changed during the backup period, it is often best to add an extra day
to the backup increment for safety.
v For each sequential backup of a directory in which only one day of data is
backed up, it might be necessary to perform many restores of many backups to
restore the entire directory from scratch. For instance, if the directory was
backed up in its entirely on the first of the month, and only a single day of
changes was backed up each following day, it will take a restore from 22
different tapes to restore the entire directory on the 22nd day of the month.
Unlike all other backup types, this type of backup contains no Logical Volume
Manager (LVM) information. Therefore, you cannot use this backup, regardless of
its contents, to recreate volume groups, logical volumes, or filesystems. Using this
backup option for backing up all files on the system (from the / directory) does not
provide a backup that can be used to reinstall the system.
It is also important to note that some files and directories on the system, such as
certain contents of the root (/) and /usr filesystems, should never be restored from
a backup, as this would adversely affect the active system configuration and can
cause either system errors or a complete system failure. Use this option only to
back up files and directories that contain non-system-related information, such as
user data and application programs.
To back up files or directories:
1. From the Backup Options menu, select Backup Files or Directories.
Note: From a command line, type smit sb_mkdirback.
2. On the Device Selector screen, highlight the device you want to use and press
Enter.
3. One of the following screens is displayed. These screens show the default
options and values available when you back up to either a local tape drive or a
disk image file.
4-24
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Backup Files or Directories
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
Hostname of server
Device name
Report output type
Rewind tape before starting backup?
Forward to End of Tape Before Starting Backup
Compress data before writing to media?
Backup only files changed within #days (0=all)
User description
Buffer size (in Kbytes)
Device name for remote volume prompt
* File(s) or starting directory(s)
[MORE]
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
[Entry Fields]
hambone.austin.ibm.com
/dev/rmt0
progress indicator
yes
no
no
[0]
[]
[64]
[]
[/tivoli/TEC /home/gues>
F3=Cancel
F7=Edit
Enter=Do
+
+
+
+
#
+#
+
F4=List
F8=Image
Figure 4-10. Default Options and Values Available When Backing Up to Tape
Backup Files or Directories
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
Hostname of server
Images directory
* Backup file ID
Dynamically Created Backup File ID
Report output type
Compress data before writing to media?
Backup only files changed within #days (0=all)
User description
Host read permission
User read permission
Buffer size (in Kbytes)
Overwrite existing backup with same ID?
* File(s) or starting directory(s)
[MORE]
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
F3=Cancel
F7=Edit
Enter=Do
[Entry Fields]
hambone.austin.ibm.com
/sysback.images/fatrat>
[03071041]
no
progress indicator
yes
[0]
[]
all hosts
all users
[64]
no
[/tivoli/TEC /home/gues>
+
+
+
#
+
+
+#
+
+
F4=List
F8=Image
Figure 4-11. Default Options and Values Available When Backing Up to Disk Image File
The fields have the following definitions:
Hostname of server:
The hostname of the data destination system on which the
selected backup device or file exists. You cannot change this
field. To select a different hostname, press F3 to return to the
menu and reselect this option to obtain a list of hosts and
device options.
Chapter 4. Performing Backups
4-25
Device name
The device, if a tape drive or virtual device was selected. You
cannot change this field from this screen.
Images directory
The directory name, if previously selected. You cannot change
this field from this screen.
Backup file ID
Default=date/time (MMDDhhmm). Uniquely identifies the
backup. You can change this field to any value but it cannot
contain single (’) or double (″) quotes. This field is active only
when you back up to a disk image file.
Dynamically Created Backup File ID
Generates a dynamic backup file ID. This option uses the same
current date and time values at the time of command execution
as the ID. If a dynamic backup file ID already exists and this
option is set to “yes”, the existing ID is overwritten.
Report output type
Default=progress indicator. Keep the default value if you want
to display a progress indicator during the backup process,
which indicates the approximate total backup time and amount
completed. Press Tab to select “file list” if you want to display
a list of files as they are being backed up, or Tab once more to
show “only errors” that occur during the backup.
Compress data before writing to media?
Default=“no” for tapes and “yes” for disk image backups. Press
Tab to change the option. “yes” indicates that the data will be
compressed before it is written to the backup media.
Compressing the data usually increases backup performance
and uses between 25 and 40% less space on the media. For best
performance, do not select this option if the backup device
provides its own hardware data compression. Compressing
data with software uses considerably more CPU cycles to
process the data.
Backup only files changed since #days (0=all)
Default=0. To backup only the files that have changed within a
certain number of days, enter the number of days here. If, for
instance, your last file or directory backup was at this time
yesterday, enter “1” to backup all files changed since that time.
Note that a day is considered exactly 24 hours, so if your
backup was earlier in the day yesterday than it was today, you
should backup at least two days of changes.
User description
Default=(Blank/None). Enter up to 60 characters of descriptive
information to add to the volume label on the backup. Do not
use single (’) or double (″) quotation marks in the description.
Host read permission
Default=all hosts. By default, all hosts with access to the system
on which this file is written have permission to read, or restore
from, this image. Press Tab to limit read permission to the
original host or the server on which the file is written. This
option is displayed only when you back up to a disk image file.
4-26
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
User read permission
Default=all users. By default, any user on a host with access to
the system and directory on which this file is written has
permission to read, or restore from, this image. Press Tab to
limit the read permission to only the original user, the root user
on the original host, or the server to which the file is written.
This option is displayed only when you back up to a disk
image file.
Rewind tape before starting backup?
Default=yes. Press Enter to rewind the tape before the backup
begins. If you are placing multiple backup images onto a single
tape, then select “no” to place this backup on the tape at the
current location. This option is displayed only when you back
up to a tape device. For a system backup, tape boot information
is placed on the tape at the beginning of the first backup. This
information is not placed on the tape for subsequent backups.
Note: If this is the first backup to this tape, or if you have
restarted the system, reinserted the tape, or changed the
tape drive block size, you should select to rewind the
tape. Only if you completed another backup and have
not performed any of these functions will the tape be
positioned correctly at the end of the previous backup
image.
Forward to End of Tape Before Starting Backup
Forwards to the end of the last SysBack backup on the tape for
all backup types. Use this option when stacking multiple
backups on a tape that has been rewound or ejected. Also use
this option when the system has been restarted or when the
block size of the tape driver has been changed from the last
backup.
Buffer size (In Kbytes)
Default=64. Changing this value either increases or decreases
the amount of data that is written to the output device in a
single I/O operation. Some devices with small buffers might
require you to reduce this value, while other devices with very
large buffers can benefit from increased performance by
increasing this value. If you are unsure, use the default value.
This value must be a multiple of the tape block size.
Device name for remote volume prompt
Default=(Blank/None). Enter a TTY, LFT or PTS device name to
send the volume prompt to a specified device rather than to the
current SMIT screen. If you are backing up to a remote host,
the device name you specify is attached to the remote host.
This field is displayed only when the backup device is a tape
drive or virtual device.
Examples of device names are /dev/tty0, /dev/lft0 and /dev/pts/5.
You can determine the terminal device name by typing tty at
the command line on that device.
Overwrite existing backup with same ID?
Default=no. If the Backup File ID field contains a value already
used for an existing system backup, change this field to “yes”
by pressing Tab. Doing so indicates that you want to overwrite
Chapter 4. Performing Backups
4-27
the previous backup with this backup. After you begin this
backup, the prior disk image files are removed and cannot be
recovered. This field is displayed only when you back up to a
disk image file.
File(s) or starting directory(s)
Default=blank. Enter the starting directory where you want the
backup to begin. You can enter multiple directory names if you
want to back up all of these directories in a single backup. If
you want to back up the entire contents of a filesystem, specify
the mount point for that filesystem (for example, /home). Press
F4 to obtain a list of all currently mounted local filesystems and
use F7 to select the starting directory from the list.
Directory names must be in one of the following formats:
v /directory/filename
v /directory
v ./filename
v ./directory/filename
v ./directory
All files will be backed up relative to the root (/) directory by
automatically inserting a period (.) at the beginning of each
filename, if it does not already exist.
4. After confirming the above entries, make sure the backup media is properly
inserted in the drive and press Enter to begin the backup.
Note: If you are backing up to a tape drive, the tape will not be rewound upon
completion of the backup. This enables you to perform additional backups
(also without rewinding), thereby “stacking” the backups onto a single tape.
4-28
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Chapter 5. Listing or Verifying Backups
This chapter describes the steps for listing or verifying the contents of a SysBack
backup.
Listing Backup Contents
Use this option to list the contents of any SysBack backup. This option shows the
header information for the backup, a list of the logical volumes or filesystems
included (except on file/directory backups), and optionally a complete list of files
and directories on the backup.
To
1.
2.
3.
4.
list the contents of a backup:
At a command line, type smit.
From the SMIT menu, select System Backup & Recovery for AIX.
Select Backup & Recovery Options.
Select List Backup Contents.
Note: You can also access this menu using the fastpath. To do this, type smit
sb_list at a command line.
5. On the Device Selector screen, highlight the device you want to use and press
Enter.
6. The following screen is displayed. This screen shows the default options and
values available when you list the contents of a backup to a local tape drive.
List Backup Contents
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
[Entry Fields]
hambone.austin.ibm.com
/dev/rmt0
[1]
yes
Hostname of server
Device or file name
Backup sequence number
Display file list?
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
F3=Cancel
F7=Edit
Enter=Do
+
F4=List
F8=Image
Figure 5-1. Default Options and Values Available When Listing the Contents of a Backup from a Local Tape Drive
The fields have the following definitions:
Hostname of server
© Copyright IBM Corp. 2002
The server hostname if a server option was
selected on the prior device selector screen. You
cannot change this field.
5-1
Device name
The device, if a tape drive or virtual device was
selected. You cannot change this field from this
screen.
Backup sequence number
Default=1. If the multiple backups were
performed sequentially to the device, select the
sequence number of the backup to list files for.
This field is displayed only when you are
listing the contents of a backup from a tape
drive or virtual device.
Default=yes. To display the list of files on the
backup in addition to the backup header
information, select “yes.” You can press Tab to
change the value to “no.” This field has no
affect when you are listing the contents of a
logical volume backup, because this backup
does not contain individual files.
7. Press Enter when all fields are correct.
Display file list?
The system begins reading the tape. If you selected a backup sequence number
other than “1,” the tape first forwards past the prior backups before reading the
backup information. The contents of the backup are displayed, along with a list of
logical volumes and filesystems, and the optional list of files and directories on the
backup.
The following is an example list of the backup header and images (filesystems and
logical volumes) information:
Backup Header
------------Date:
Sun Mar 9 13:48:02 1997
User:
root
Host:
sbclient
Type:
TOC - Volume Group
Data:
vg00
Size:
33 megabytes
Packed: N
Buffer: 64 Kbytes
Bkend: B
Volume: 1
Images
Image
----1
2
3
contained on this backup:
VG name
LV Name
Size
Mount Point
-------------- -------------- ----- ----------------vg00
lv02
1
/testfsnew
vg00
rawlv2
24
vg00
rawlv3
8
-
The list above indicates that this is a volume group backup of the vg00 volume
group and includes “raw” logical volumes rawlv2 and rawlv3 and the /testfsnew
filesystem.
Verifying Files on System Backup
It is often a good idea to verify all of the data on a backup to ensure that the
backup was written properly. For backups containing multiple data options, such
as multiple volume groups or multiple filesystems, you can verify all or part of the
backup. For instance, if you created a system backup containing multiple volume
groups, you can verify one volume group, multiple volume groups, or all volume
5-2
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
groups on the backup. Likewise, for a logical volume backup containing multiple
logical volumes, you can verify one or more logical volumes.
To verify data on a backup:
1. From the Backup & Recovery Options menu, select Verify a Backup.
Note: From a command line, type smit sb_verify.
2. On the Device Selector screen, highlight the device you want to use and press
Enter.
3. If you selected a tape drive or virtual device, enter the backup sequence
number. The default value is “1,” indicating that the backup to be verified is
the first backup on the media. If you stacked multiple sequential backups on
the media and want to verify a different backup, type the backup number and
press Enter.
4. Next, the backup is read to determine the backup type. If the backup is not a
file or directory backup, you are asked to select the data to be verified. A list of
verify options appear, dependent on the backup type. For instance, if this is a
volume group backup, the list includes volume groups to verify. If this is a
filesystem backup, the list includes filesystems. To select a single option to
verify, highlight the option and press Enter. To select multiple options, highlight
each line and press F7. When all selections have been made, press Enter to
continue.
5. The following screen shows the default options and values that appear when
you verify the contents of a volume group backup of vg00 from a local tape
drive:
Verify a Backup
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
[Entry Fields]
hambone.austin.ibm.com
/dev/rmt0
Volume_Group
both
[]
[rootvg]
Hostname of server
Device name
Verify data type
Report output type
Device name for remote volume prompt
Data to verify (of above type)
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
F3=Cancel
F7=Edit
Enter=Do
+
F4=List
F8=Image
Figure 5-2. Default Options and Values Available When Verifying the Contents of a Volume Group Backup
The fields have the following definitions:
Hostname of server
The server hostname if a server option was
selected on the prior device selector screen. You
cannot change this field.
Chapter 5. Listing or Verifying Backups
5-3
Device name
The device, if a tape drive or virtual device was
selected. You cannot change this field from this
screen.
Disk image file
The name of the disk image file previously
selected. This field is displayed only if a disk
image file name was selected. You cannot be
change this field from this screen.
Verify Data Type
The backup type. It indicates the type of data,
as shown in the Data to verify field, that is to
be verified. This field is for information only
and cannot be changed.
Report output type
Default=progress indicator. Keep the default
value if you want to display a progress
indicator during the backup process, which
indicates the approximate total backup time
and amount completed. Press Tab to select “file
list” to display a list of files as they are being
backed up, or Tab once more to show “only
errors” that occur during the backup.
Device name for remote volume prompt
Default=(Blank/None). Enter a TTY, LFT or PTS
device name to send the volume prompt to a
specified device rather than to the current SMIT
screen. If you are backing up to a remote host,
the device name you specify is attached to the
remote host. This field is displayed when the
backup device is a tape drive or virtual device.
Examples of device names are /dev/tty0,
/dev/lft0 and /dev/pts/5. You can determine
the terminal device name by typing tty at the
command line on that device.
Data to verify (of above type) Unless the backup is a file/directory Backup,
this field contains the data to verify as selected
on the prior screen. This can be a list of volume
groups, filesystem, or logical volumes,
depending on the type of backup to be verified.
For a file/directory backup, this field is left
blank and any contents are ignored, because the
entire contents of the backup must be verified.
6. Press Enter when all fields are correct.
Each data type specified is read in its entirety to ensure the data is readable on the
media. Either a list of files or progress indicator is displayed during the
verification as selected above.
5-4
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Chapter 6. Recreating or Restoring from Backups
There are many reasons to restore data from a backup. The data might have been
corrupted due to improper user input or application errors, or the data might be
inaccessible due to various hardware problems.
If data was inadvertently deleted or corrupted by a user or application, you will
probably need to restore the data from the backup. If a hardware failure occurred,
you need to perform additional steps before you can restore the data. For instance,
if a disk failure occurs that makes a volume group inaccessible, it will probably be
necessary to delete and recreate the volume group, recreate the filesystems and
logical volumes, and mount the filesystems, before the data can be restored. Any
SysBack backup type can be used in this process, but the backup must contain the
desired data, as well as the desired system information, to recreate or restore the
desired data.
Note: This section does not describe the process for reinstalling the system, only
recreating and restoring data onto an already active system. For information
on reinstalling the system from a system backup, refer to Chapter 16,
“System Installation and Maintenance”, on page 16-1.
Removing Volume Groups, Logical Volumes, and Filesystems
Before any volume group, logical volume, or filesystem can be recreated on an
active system, you must remove the old volume group, logical volume, or
filesystem. This process is not detailed within this manual because the steps vary
too widely depending on the reason for the recreation process. For detailed
assistance on any of these activities, contact your preferred AIX Technical Support
Organization. The following guidelines might help in deleting the old system
information:
v Volume Group: To delete a volume group, first refer to the information below
on how to make filesystems and logical volumes inactive. Then, use the AIX
varyoffvg command to make the volume group inactive and the exportvg
command to remove the volume group information from the system
configuration database.
v Filesystems: To delete a filesystem, the filesystem must be unmounted, even if
the filesystem is currently inaccessible. You can do this with the AIX umount
command. To use this command, no user can currently be changed to (cd
command) any directory within the filesystem, and no process on the system can
have any file in the filesystem open.
If you are removing the volume group containing this filesystem, the filesystem
will be removed when the volume group is exported. However, to remove only
a select filesystem, first unmount the filesystem, then use the AIX rmfs
command to remove it.
v Logical Volumes: To delete a logical volume, no process can have the logical
volume open. Certain logical volumes used by the system are normally open by
system processes, and the steps to make them inactive vary depending on the
logical volume type. To list the logical volumes for a volume group, use the lsvg
-l VGname command. This command also shows the logical volume type and
whether the logical volume is currently opened (active).
© Copyright IBM Corp. 2002
6-1
The following information should be helpful for making the logical volume
inactive based on its type:
jfs
This is the default logical volume. jfs indicates the logical
volume is used for a journaled filesystem. If this is the case, refer
to the instructions above for Filesystems. Otherwise, an
unknown process on the system has the logical volume open.
jfslog
This is a logical volume used by filesystems in the volume
group. This logical volume is made inactive automatically when
all of the filesystems that reference it are unmounted. View the
/etc/filesystems file to see which filesystems reference which
jfslog logical volumes.
paging
This logical volume is a paging space. An active paging space
cannot be disabled as long as the system is running. Instead, you
must deactivate the paging space for the next system boot using
the command chps -an LVname. After doing so, the system
must be rebooted for the paging space to be inactive.
dump
This logical volume is used for a system crash dump and is
referred to as the dump device. You must disable the system
dump to make this logical volume inactive. To do so, execute the
command sysdumpdev -Pp /dev/sysdumpnull and sysdumpdev
-Ps /dev/sysdumpnull.
Any other logical volume types are user-defined and have no specific meaning to
the system.
If you are removing the entire volume group containing the logical volumes, the
logical volumes are removed along with the volume group information. If,
however, you are only removing the logical volumes, use the rmlv command to
remove the logical volumes once they are inactive.
Recreating Volume Groups, Logical Volumes, and Filesystems
If you experience a hardware failure that requires you to recreate a volume group,
logical volume, or filesystem, you can use either the system or volume group
backup. A filesystem backup can also be used to recreate filesystems while a
logical volume backup can be used to recreate logical volumes. These “containers”
must be recreated to provide a place to restore the data.
You can recreate one or more volume groups, logical volumes or filesystems as
they are defined on the backup, or you can optionally change the volume group,
logical volume or filesystem characteristics, including the disk location, filesystem
and logical volume sizes, or any other attribute.
Note: The containers you want to recreate must have information pertaining to
them on the backup media. A volume group backup of only the vg00
volume group cannot be used to recreate any other volume group. However,
this backup can be used to recreate single logical volumes or filesystems that
were contained within volume group vg00. The system backup, however,
always contains information about all volume groups, logical volumes, and
filesystems, even if not all volume group data was included on the backup.
6-2
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
You do not have to use the backup media to recreate a volume group, logical
volume or filesystem on the same system. You can use the media to perform the
recreation on another system, thereby copying an environment from one system to
another.
Note that this option recreates the volume group, logical volumes, and filesystems,
but cannot restore the data. You can use the Restore Data from a Backup option to
restore data.
To recreate one or more volume groups, logical volumes, or filesystems:
1. At a command line, type smit.
2. From the SMIT menu, select ∏.
3. Select Backup & Recovery Options.
4. Select Recreate Volume Groups, Logical Volumes or Filesystems.
Note: From a command line, type smit sb_create.
5. On the Device Selector screen, highlight the device you want to use and press
Enter.
6. If you selected a tape drive or virtual device, enter the backup sequence
number. The default value is “1”, indicating that you want to use the first
backup on the media. If you “stacked” multiple sequential backups on the
media, and want to use the information from a different backup, enter the
backup number and press Enter.
7. If the backup is a system or volume group backup, select the type of backup
you are going to recreate. You have two options:
Volume Group (all LVs and filesystems)
Logical Volume and/or Filesystem
Highlight the type of backup you are creating and press Enter.
If the backup is a Filesystem or Logical Volume Backup, the recreate type is
assumed to be “Logical Volume and/or Filesystem.”
8. Depending on your response to the previous prompt, select from the
displayed list either a volume group or logical volume to create. To select a
single option, highlight the option and press Enter. To select multiple options,
highlight each option and press F7 to select. When you have made all
selections, press Enter to continue.
9. One of the following screens is displayed. These screens are examples of the
screens that are displayed when you recreate either a single volume group or
two logical volumes from a tape device:
Chapter 6. Recreating or Restoring from Backups
6-3
Recreate a Volume Group from a Backup
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
Hostname of server
Device or file name
Edit the Volume Group info before proceeding?
New Volume Group Name
Volume Group(s) to create
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
F3=Cancel
F7=Edit
Enter=Do
[Entry Fields]
hambone
/sysback.images/chukra>
yes
+
[]
[vg00]
F4=List
F8=Image
Figure 6-1. Recreating a Single Volume Group
Recreate a Logical Volume or Filesystem from a Backup
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
Hostname of server
Device or file name
Edit the Logical Volume info before proceeding?
New Volume Group Name
Logical Volume(s) to create
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
F3=Cancel
F7=Edit
Enter=Do
[Entry Fields]
hambone
/sysback.images/chukra>
yes
+
[]
[lv01]
F4=List
F8=Image
Figure 6-2. Recreating a Two Logical Volumes from a Tape Device
The following is an explanation of each field:
Hostname of server
The server hostname if a server option was selected on the prior
device selector screen. You cannot change this field.
Device or file name
The device or disk image file name previously selected. You cannot
change this field on this screen.
Edit the Logical Volume info before proceeding?
Default=yes. This prompt is displayed only when selecting to recreate
6-4
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
logical volumes or filesystems. If you want to view or change the
logical volume or filesystem information, such as the disk locations,
sizes, or other attributes, before recreating the logical volumes and
filesystems, keep the default value of “yes.” Otherwise, to recreate the
logical volumes or filesystems without viewing or changing the
information, press Tab to change the value to “no.”
Edit the Volume Group info before proceeding?
Default=yes. This prompt is displayed only when selecting to recreate
volume groups. If you want to view or change the volume group,
logical volume, or filesystem information, such as the disk locations,
sizes, or other attributes, before recreating the volume group and its
logical volumes and filesystems, keep the default value of “yes.”
Otherwise, to recreate the volume group, logical volumes or
filesystems without viewing or changing the information, press Tab to
change the value to “no.”
New Volume Group Name
Default=(blank/none). If you are recreating a volume group, you can
enter a new volume group name in this field. This is particularly
useful if you want to recreate a volume group but the original volume
group by the same name still exists on the system. You can not enter a
value in this field if you are creating more than one volume group at
one time.
If you are recreating one or more logical volumes, entering a value in
this field indicates that the logical volumes to be created will be
placed in the specified volume group. In this case, the volume group
name you enter must be an active volume group on the system.
Logical Volume(s) to create
This field is displayed only when selecting to recreate logical volumes
and filesystems. This field will contain the names of the logical
volumes selected on the prior selection screen. You can change the
logical volumes to create by adding or removing entries from this list,
separating each entry by a space.
Although only the logical volume name is indicated here, if the logical
volume contains a filesystem, the filesystem will be recreated.
Note that information regarding the logical volumes to recreate must
be contained on the backup media previously selected.
Volume Group(s) to create
This field is displayed only when selecting to recreate volume groups.
This field will contain the names of the volume groups selected on the
prior selection screen. You can change the volume groups to create by
adding or removing entries from this list, separating each entry by a
space. Note that information regarding the volume groups to recreate
must be contained on the backup media previously selected.
10. When you have made all selections, press Enter to begin.
The media is read to obtain the information on the volume groups, logical volumes
or filesystems selected.
If you selected to edit the volume group, logical volume or filesystem information,
a screen similar to the following is displayed:
Chapter 6. Recreating or Restoring from Backups
6-5
+----------------------------------------------------------+
|
Change Volume Group & Logical Volume Information
|
+----------------------------------------------------------+
Select Physical Volumes for Volume Groups
Change Volume Group Attributes
Select Physical Volumes for Logical Volumes
Change Logical Volume Attributes
Change Filesystem Attributes
+------------------------------------------------------------------------------+
| Select this option to change the physical volumes which are assigned to each |
| volume group.
|
+------------------------------------------------------------------------------+
| Use ARROW or Tab keys to move and ENTER to Select. Press ESC for Main Menu. |
+------------------------------------------------------------------------------+
Figure 6-3. Editing the Volume Group, Logical Volume, or Filesystem Information
The options for volume groups are not displayed when you select to recreate
logical volumes. Also, if this is a power backup, the option for changing filesystem
information is not displayed. These menus provide options for changing virtually
all attributes for volume groups, logical volumes, and filesystems. This process is
identical to reinstalling a system from a system backup. In both cases, the detailed
steps for changing this information is provided in Chapter 18, “Changing the
Volume Group, Logical Volume and Filesystem Attributes”, on page 18-1.
The volume group, logical volume, and filesystems are then checked to ensure that
they can be recreated based on the current system resources available. If there are
inconsistencies, such as not enough disk space or unassigned physical volumes
(disks), the appropriate messages are displayed and you must change the volume
group or logical volume information before the process can continue.
If there are no inconsistencies with the new system, you are asked:
Are you sure you wish to create the volume group(s) (y/n)?
or
Are you sure you wish to create the logical volume(s) (y/n)?
Answering “y” to this question recreates the volume groups or logical volumes as
specified. When the process has completed, you are asked to press Enter to return
to the SMIT screen.
Understanding Incremental Restores
If you created incremental backups, then you must understand the process of
restoring data from these backups. Examples of planning and performing
incremental backups and their corresponding restoration processes are detailed in
“Understanding Incremental Backups” on page 4-1.
6-6
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
An incremental restore requires restoring a volume group or filesystem to the state
of its last full (level 0) backup, and then reapplying each subsequent incremental
backup level until the volume group or filesystem is returned to its current state.
Because only a level 0 backup contains all of the data in either the volume group or
filesystem, this backup must be restored before any additional incremental levels
can be applied.
Restoring an incremental backup does not only add or change files in a filesystem,
but can also remove files from a filesystem. If a file was removed from the
filesystem between the time a level 0 and a level 1 backup was created, the file can
be restored during a level 0 restore, and removed from the system again during a
level 1 restore.
SysBack does not check the order in which the incremental levels are restored. The
user must ensure that the incremental backups are restored in the proper order.
SysBack requires that the user explicitly indicate that they want to restore an
incremental backup. If an attempt is made to restore a volume group or filesystem
from an incremental backup without the user indicating this intent, the restore
process does not proceed.
Restoring level 0: When a level 0 backup is restored, the filesystems on the system
that are to be restored are first cleared of all files. Then, the backup of the
filesystems is restored. At this point, the filesystems have been returned to the
exact state at which the level 0 backup was made.
Restoring other levels: After the level 0 backup is applied, the subsequent backup
level or levels must be applied to return the filesystems to their most recent state.
The number of levels to restore differs depending on the design of the incremental
backups. The important thing to remember is that the most recent of each
subsequent level must be applied in the correct order. If, for instance, you
performed multiple level 3 backups, it is only the most recent level 3 that needs to
be restored, because the most recent level 3 backup supersedes all other level 3
backups.
Before a level (other than 0) is restored, any files that currently exist in the
filesystems, but did not exist at the time the backup level was created, are removed
from the filesystems.
Restoring individual files or directories: You can restore individual files or
directories from any incremental backup level. This is not considered an
incremental restore, so it is not necessary to restore a level 0 backup first. If you
want to restore the most recent copy of a file, it might be necessary to list the files
on each backup level, in reverse order, before finding the file. The file is always
included on the level 0 backup (if it existed at that time), but later revisions of the
file can be included on any subsequent backup level.
Restoring Data from a Backup
Use the Restore Data from a Backup option to restore one or more volume
groups, filesystems, logical volumes, directories, or regular files from any type of
backup created with SysBack.
To restore from a backup:
1. From the Backup & Recovery Options menu, select Restore Data from a
Backup.
Chapter 6. Recreating or Restoring from Backups
6-7
Note: From the command line, type smit sb_restore.
2. On the Device Selector screen, highlight the device you want to use and press
Enter.
3. If you selected a tape drive or virtual device, enter the backup sequence
number. The default value is “1,” indicating that you want to restore data from
the first backup on the media. If you “stacked” multiple sequential backups on
the media, and want to use the information from a different backup, enter the
backup number and press Enter.
The backup media is then read to determine the backup type.
4. Select the type of data to be restored from the list displayed. The possible
choices include:
v Volume Group (all LVs and filesystems)
v Filesystem
v All Directories and Files
v Directory (all files within)
v Regular File
Only the options that are valid for the type of backup you are restoring from
are displayed. Highlight the type of data you want to restore and press Enter.
5. If you selected volume group, logical volume or filesystem, select the data to be
restored from the list of volume groups, logical volumes, or filesystem mount
points. To select a single option, highlight the option and press Enter. To select
multiple options, highlight each option and press F7. When you have made all
selections, press Enter to continue.
6. At the “Do you wish to list select files to restore?” prompt, select “yes” for a
list of all files contained on the backup media. You can also narrow the list by
using a search word or wildcard character. The search word can include the
wildcard (*) character. SysBack supports the use of BRE (Basic Regular
Expression) wild cards to restore a wildcard match of files. The filenames and
the wildcard must be enclosed in double quotation marks (″) to avoid
expansion by the user’s shell.
For example, the search word ″/home/j*″ indicates to restore all directories and
files that start with ″/home/j″ such as /home/john or /home/jenn. Another
example is the search word ″ab*d″, which would display the following
filenames:
/tmp/abcd
/var/spool/abduct
/home/data/aboyandhisdog
SMIT limitations prevent a list of files longer than 32768 lines from being
displayed. Therefore, the list, if longer, is truncated to this size.
If you want to display a file that contains a list of files to restore, select “no”
7. The following figures show the default options and values that are displayed
when you restore a filesystem from a volume group Backup, and regular files
from a filesystem backup:
6-8
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Restore Data from a Backup
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
[Entry Fields]
hambone
/sysback.images/chukra>
Filesystem
both
+
no
+
no
+
[]
[/fixes]
Hostname of server
Disk image file
Restore data type
Report output type
Allow restore of system data?
Allow restore of incremental backup?
Destination (if different)
Data to restore (of above type)
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
F3=Cancel
F7=Edit
Enter=Do
F4=List
F8=Image
Figure 6-4. Default Values When Restoring a Filesystem from a Volume Group Backup
Restore Data from a Backup
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
Hostname of server
Disk image file
Restore data type
Report output type
Allow restore of system data?
Allow restore of incremental backup?
Destination (if different)
File Containing list of data to be Restored
Data to restore (of above type)
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
F3=Cancel
F7=Edit
Enter=Do
[Entry Fields]
hambone
/sysback.images/chukra>
Regular_File
both
+
no
+
no
+
[]
[]
[/backups/apache.pdf /b>
F4=List
F8=Image
Figure 6-5. Default Values When Restoring a File from a Filesystem Backup
The following is an explanation of each field:
Hostname of server
The server hostname if a server option was selected on the prior device
selector screen. You cannot change this field.
Device name
The device, if a tape drive or virtual device was selected. You cannot
change this field from this screen.
Chapter 6. Recreating or Restoring from Backups
6-9
Disk image file
The name of the disk image file previously selected. This field is
displayed only if a disk image file name was selected. You cannot
change this field from this screen.
Restore data type
The type of data to be restored. This field is for information only and
cannot be changed.
Report Output type
Default=progress indicator. Keep the default value if you want to
display a progress indicator during the backup process, which indicates
the approximate total backup time and amount completed. Press Tab to
select “file list” to display a list of files as they are being backed up, or
Tab once more to show “only errors” that occur during the backup.
Device name for remote volume prompt
Default=(Blank/None). Enter a TTY, LFT or PTS device name to send
the volume prompt to a specified device rather than to the current
SMIT screen. If you are backing up to a remote host, the device name
you specify is attached to the remote host. This field is displayed only
when the backup device is a tape drive or virtual device.
Examples of device names are /dev/tty0, /dev/lft0 and /dev/pts/5.
You can determine the terminal device name by typing tty at the
command line on that device.
Allow restore of system data?
Default=no. If this field is set to “no”, the restore process cannot restore
certain directories, logical volumes, or filesystems that might affect the
system operation. Examples are the hd4 (/ filesystem) logical volume,
the /usr filesystem, or the /etc/objrepos directory. Press Tab to change
this field to “yes” if you want to allow any system data to be restored,
but be aware that this might cause system errors or a complete system
failure depending on the inconsistency of the data being restored and
the current system configuration.
Allow restore of incremental backup?
Default=no. Change this field to “yes” if you are restoring entire
filesystems or volume groups from an incremental backup. The
restoration of incremental data can have a different effect than restoring
regular backup data. Information on restoring incremental backups is
detailed in “Understanding Incremental Restores” on page 6-6.
An error message is displayed and the process terminates if you
attempt to restore data from an incremental backup without setting this
field to “yes.”
Destination (if different)
Default = blank. If you are restoring data to its original location, leave
this field blank. To select an alternate “relative” logical volume or
directory for the restored data, select one of the following:
v For a file/directory backup, enter a new directory. If the file to be
restored from the media is /home/tony and you want to restore the
file to /tmp directory, the file is restored “relative” to /tmp, resulting
in the filename /tmp/home/tony.
v For a filesystem backup, enter a new directory. The contents of the
entire filesystem will move to the new directory. The new filenames
will not be relative to the previous filesystem mount point. For
6-10
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
instance, the /home/tony filesystem (containing the file
/home/tony/file1) can be moved to the /test directory; the resulting
file is /test/file1 (the /home/tony directory prefix is removed).
v For logical volume backups, enter a new logical volume name. The
data is restored to the new logical volume. An “end of file” error
occurs when writing the data if the new logical volume is smaller
than the original.
Note: You cannot specify a new destination when you select
multiple logical volumes, volume groups, or filesystems to
restore.
File Containing a List of Data to be Restored
Specify the fully qualified path name to a file that contains the list of
files to be restored from this backup image. Use this option when
restoring large numbers of files.
Data to restore (of above type)
Unless the backup is a file/directory backup and you select to restore
all files and directories, this field contains the data to restore as selected
on the prior selection screen. This can be a list of volume groups,
filesystems, logical volumes, directories or regular files, depending on
the type of backup to be restored. For a file/directory backup, leave
this field blank if you want to restore all files.
8. Press Enter when all fields have been entered correctly. Either a progress
indicator or list of files, if selected, is displayed as the data is read from the
media and selected data is restored.
Chapter 6. Recreating or Restoring from Backups
6-11
6-12
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Chapter 7. Scheduled Backups and Scripts
The SysBack Scripts and Scheduling functions provide a mechanism to automate
common backup functions that require little or no user interaction. The SysBack
Scripts functions help you build command syntax into a Korn Shell script that can
be scheduled with SysBack. The SysBack Scheduling functions enable you to take
these scripts and schedule them to run automatically at a specified time or interval.
The Scheduling functions act as a front-end to crontab and enables you to add,
update, and remove SysBack backup scripts to crontab.
Scripts
Accessing the Backup Scripts Menu
The Backup Scripts menus enable you to create, change, or remove backup scripts.
To access the Backup Scripts menus:
1. From the SysBack Main Menu, select Backup Schedules and Scripts.
Note: At a command line, type smit sb_script_sched.
2. Select Backup Scripts.
Note: At a command line, type smit sb_script.
Creating a Backup Script File
When you create the backup script file, the SMIT options with which you are
presented give the appearance that a backup will be performed. These options are
only used to create the backup command sequence, and not to start a backup.
The scripts create with the Create a Backup Script File option are stored in the
/usr/lpp/sysback/sbscripts directory, unless a fully qualified path name is specified
in the script name field.
Create a Backup Script File
Move cursor to desired item and press Enter.
Create
Create
Create
Create
Create
F1=Help
F5=Reset
F9=Shell
a
a
a
a
a
System Backup Script
Volume Group Backup Script
Filesystem Backup Script
Logical Volume Backup Script
File/Directory Backup Script
F2=Refresh
F6=Command
F10=Exit
F3=Cancel
F7=Edit
Enter=Do
F4=List
F8=Image
Figure 7-1. Create a Backup Script File Menu Options
To create a backup script file:
1. From the Backup Scripts Menu, select Create a Backup Script.
© Copyright IBM Corp. 2002
7-1
Note: At a command line, type smit sb_script_c.
2. Select the type of backup script you want to create. The following choices are
available:
v Create a System Backup Script
v Create a Volume Group Backup Script
v Create a Filesystem Backup Script
v Create a Logical Volume Backup Script
v Create a File/Directory Backup Script
Note: For this example, Create a System Backup Script is used.
3. On the Device Selector screen, highlight the device you want to use and press
Enter.
4. The following screen is displayed.
Create a System Backup Script
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
[TOP]
Backup Script Name
Description
Overwrite Existing Script
HOSTNAME of server
Device Name
Create a Power Backup?
Report output type
Platform/kernel type for tape boot image
Network install support to include
Compress data before writing to media?
User Description
Buffer size (in Kbytes)
Preserve Physical partition mapping?
Device name for remote volume prompt
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
[system_backup]
[System Backup]
No
[/dev/rmt0]
No
File List
rspc/MP
No
[]
[64]
No
[]
F3=Cancel
F7=Edit
Enter=Do
F4=List
F8=Image
Figure 7-2. The Create a System Backup Script Menu
The first three options are described below. The other options are identical to
the options available when you create a system backup. For descriptions of
those options, see the instructions for creating system backups in Chapter 4,
“Performing Backups”, on page 4-1.
The first three options have the following definitions:
Backup Script Name (Required)
The fully-qualified filename for the backup script. If a directory is not
specified, the script is created in the /usr/lpp/sysback/sbscripts
directory by default.
Description (Optional)
The purpose or content of the script. If no description is provided, the
description, by default, reflects the type of backup performed by the
script.
You can use any character or number, except for single (’) or double (″)
quotation marks.
7-2
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Overwrite Existing Script (Optional)
Determines whether or not to overwrite an existing backup script. The
default is “No”
5. Press Enter to confirm your selections.
The following file is an example of a backup script generated using the options
specified above.
#!/bin/ksh
# Command filename: /usr/lpp/sysback/scripts/system_backup
# Description: System Backup
sysback -v -f /dev/rmt0 -T rspc -k mp
exit $?
Note: The commands contained in the backup script file are not checked for
validity. Always verify the validity of the backup script by performing a
backup using the same options as those in the backup script.
Changing a Backup Script File
Use the Change a Backup Script File option to edit the contents of a backup script
created by SysBack. You can add additional commands or change the existing
commands and options.
The Change a Backup Script File option uses the AIX vi editor and provides help
screens and pre-programmed function keys. However, you can edit the script file
with any text editor.
To change a backup script file:
1. From the Backup Scripts menu, select Change a Backup Script File.
Note: At a command line, type smit sb_script_change.
2. Select the script file you want to edit and press Enter.
3. The following vi help screen is displayed. This screen lists the actions you can
perform, as well as the corresponding function key or key sequences.
EDITOR HELP SCREEN
You are about to enter the “vi” editor. If your terminal type
provides function keys, the following function keys may be used
for editing the file. The standard “vi” commands shown may also
be used if function keys are not provided.
FUNCT KEY
---------F1
F2
F3
F4
F5
F6
F7
F8
F9
ACTION
-------Display this help screen
Exit without saving changes
Save changes and exit
Page Back
Page Forward
Begin/End Insert Mode
Insert Line (F6 to quit)
Delete Character
Delete Line
STANDARD VI KEY
----------------:e#
:q!
:wq!
CTRL-B
CTRL-F
i/ESC
o/ESC
x
dd
=== Press F1 or “:e#” to edit file ===
Figure 7-3. The vi Editor Help Screen
Chapter 7. Scheduled Backups and Scripts
7-3
Notes:
a. Press F1 at any time to return to the help screen.
b. Press F2 at any time to exit the editor without saving the changes.
4. Press F1 to exit the help screen and begin editing the file.
Removing a Backup Script File
Use the Remove a Backup Script File option to remove existing SysBack script
files.
To remove a backup script file:
1. From the Backup Scripts menu, select Remove a Backup Script File.
Note: At a command line, type smit sb_script_remove.
2. Select the script file you want to edit and press Enter.
3. Press Enter to confirm the removal of the script file.
Accessing the Backup Schedules Menus
The Backup Schedules menus enable you to create, update, list, and remove
backup schedules.
To access the backup schedules menus:
1. From the SysBack Main Menu, select Backup Schedules and Scripts.
Note: At a command line, type smit sb_script_sched.
2. Select Backup Schedules.
Note: At a command line, type smit sb_sched.
Creating a Backup Schedule
Use the Create a Backup Schedule option to schedule an existing backup script.
This script can be one created by SysBack or a user-defined script.
The Create a Backup Schedule option provides a front-end to the AIX crontab and
features a menu-based method for creating a schedule.
To create a backup schedule:
1. From the Backup Schedules menu, select Create a Backup Schedule.
Note: At a command line, type smit sb_sched_c.
2. The following screen is displayed:
7-4
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Create a Backup Schedule
Type or select values in entry fields
Press Enter AFTER making all desired changes.
Script Name
Schedule ID
Schedule Description
Minute (0-59)
Hour (0-23)
Day (1-31)
Month (1-12)
Week Day (0-6)
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
[/usr/lpp/sysback/scripts/system_backup
[Friday and Wednesday Full System Backup}
[0]
[23]
[]
[]
[3,5]
F3=Cancel
F7=Edit
Enter=Do
F4=List
F8=Image
Figure 7-4. The Create a Backup Schedule Menu
The fields have the following definitions:
Script Name (Required)
The name of an existing script file. Press F4 for a list of script files, or
type the fully qualified path and file of the script file you want to use.
Schedule ID (No Input)
The Schedule ID is generated at command execution and is related to
the schedule. You cannot enter any information in this field.
Schedule Description (Optional)
A description of the backup schedule. You can enter any text except
single (’) or double (″) quotation marks.
Minute (Required)
Specifies the starting minute, a number between 0 and 59. Press F4 for
a list of valid values.
Hour (Required)
Specifies the starting hour, in military time. Use a number between 0
and 23, where 0 equals 12:00 AM and 23 equals 11:00 PM.
Day (optional)
Specifies the day of the month, a number between 1 and 31. Press F4
for a list of valid values. If both this field and the WeekDay field are
left blank, the schedule runs every day.
Month (optional)
Specifies the starting month, a number between 1 and 12. Press F4 for a
list of valid values. If this field is left blank, the schedule runs every
month.
WeekDay (optional)
Specifies the day of the week, a number between 0 and 6, where 0
equals Sunday. Press F4 for a list of valid values. If both this field and
the Day field are left blank, the schedule runs every day.
Important Field Input Notes:
a. Two numbers separated with a dash indicate
an inclusive range. For example, to schedule
a command to run Tuesday through Friday,
type ″2-5″ in the WeekDay field.
Chapter 7. Scheduled Backups and Scripts
7-5
b. Numbers separated with commas indicate
that the schedule should run on separate
days. For example, to run a schedule on the
first and last days of January, type ″1″ in the
Month field and ″1,31″ in the Day field.
3. Press Enter to commit the selections.
Note: Do not make changes to the schedule through manual manipulation of
crontab. Because SysBack acts as a front-end to crontab, making changes
manually causes the SysBack scheduling features and crontab to be
unsynchronized.
Updating a Backup Schedule
To update an existing backup schedule that was created by SysBack:
1. From the Backup Schedules menu, select Update a Backup Schedule.
Note: At a command line, type smit sb_sched_u.
2. Select the schedule you want to update and press Enter.
3. The following screen is displayed:
Change or Update an Existing Backup Schedule
Type or select values in entry fields
Press Enter AFTER making all desired changes.
Script Name
Schedule ID
Schedule Description
Minute (0-59)
Hour (0-23)
Day (1-31)
Month (1-12)
Week Day (0-6)
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
[/usr/lpp/sysback/scripts/system_backup]
SB1
[Friday and Wednesday Full System Backup]
[0]
[23]
[]
[]
[3,5]
F3=Cancel
F7=Edit
Enter=Do
F4=List
F8=Image
Figure 7-5. The Change or Update an Existing Backup Schedule Menu
For descriptions of the above options, see “Creating a Backup Schedule” on
page 7-4.
4. Press Enter to commit the changes.
Listing Backup Schedules
To list existing backup schedules created by SysBack, select List Backup Schedules
from the Backup Schedules menu.
Note: From a command line, type smit sb_sched_l.
The following is an example of the output produced.
Command: /usr/lpp/sysback/sbscripts/file.lasher.2
ID: SB1
Description: Sample Test Schedule for host lasher
Execution Time:
7-6
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
-------------Every: 15th in the month(s) of June at 9:25 a.m.
---------------------------
Removing a Backup Schedule
To remove an existing backup schedule:
1. From the Backup Schedules menu, select Remove a Backup Schedule.
Note: At a command line, type smit sb_sched_r.
2. Select the schedule you want to remove and press Enter.
3. Press Enter again to confirm the removal of the backup schedule.
Note: Removing the backup schedule in this manner removes the schedule from
SysBack, as well as its corresponding entry in crontab.
Chapter 7. Scheduled Backups and Scripts
7-7
7-8
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Chapter 8. Local User Access
When you install SysBack, all users on the local system are, by default, granted
access to all backup devices on the system. They are also allowed to create backups
in the default directory /usr/lpp/sysback/bf/local/all. The following options are
used to change user access by adding or deleting the devices or directories that
each user can access.
To access the local device or directory access menu:
1. At a command line, type smit.
2. Select System Backup & Recovery for AIX.
3. Select Configuration Options.
4. Select Local User Access.
You can also access this menu using the fastpath. To do this, type smit sb_local at
a command line.
The following screen is displayed:
Local User Access
Move cursor to desired item and press Enter.
Add or Change Local User Device/Directory Access
List User Access
Remove User Access
F1=Help
F9=Shell
F2=Refresh
F10=Exit
F3=Cancel
Enter=Do
F8=Image
Figure 8-1. The Local User Access Menu
Detailed instructions for each of the selections in the Local User Access menu
follow.
Understanding Local User Access Permissions
SysBack provides flexible access for device, directories, and so on. For instance,
you can assign specific devices and directories to each user on the system, or
assign a device to all users on the system except specific users. Before you assign
access, it is important to understand the priority of the permission records and
how they are used.
© Copyright IBM Corp. 2002
8-1
Assume the following records exist:
Users to allow
Users to deny
Devices
Directories
all
greg peter bobby
/dev/rmt0 vdev0
/usr/lpp/sysback/bf/local/all
root
all
usr/lpp/sysback/bf/local/root
alice
/dev/rmt1
Based on the information above, the following statements are true:
1. Users greg, peter and bobby do not have backup access on this system because
they are explicitly denied access under all users. You can only deny access to
users when defining access for “all” other users. You cannot create a permission
record for any users that have been explicitly denied access.
2. User root has access to all devices (tape and virtual). User root is also provided
a disk image file backup option, with the default filename contained in the
/usr/lpp/sysback/bf/local/root directory. You can change the default name of the
backup file at backup time, but the file must exist in the
/usr/lpp/sysback/bf/local/root directory or one of its subdirectories.
3. User alice has access only to tape device /dev/rmt1. Because there are no
directories defined for backup image files, no disk image backup option is
provided to this user.
4. All other users (all users except those explicitly denied and those explicitly
defined) are provided access to tape device /dev/rmt0 and virtual device
vdev0. Those users are also provided a disk image backup option in the
/usr/lpp/sysback/bf/local/all directory. They can back up to only those devices
and directories defined in this record.
Also note that, if you remove user greg from the list of users denied access, he is
automatically provided access under the same record as all other users. Likewise,
removing user alice does not disable access, because she is not explicitly denied,
but her permission also defaults to that of all users.
Adding or Changing Local User Device/Directory Access
All users on the system are provided access to all of the backup devices on the
system when SysBack is first installed. This is accomplished by providing a
permission record for a user called all. You might not want certain users to have
access to all devices and backup directories, so you might want to deny specific
users access, or change the device or directory list that specific users are allowed to
use. The devices and directories specified when using this option determine the list
of backup options that users receive when performing backups using SMIT, and
those devices or directories that other SysBack commands, executed at the
command line, can read or write to.
If you do not want all users on the system to be granted the same access, remove
access to user all before setting up access to specific users. This can help avoid
confusion in the future.
To add or change device or directory access to local users:
1. From the Local User Access menu, select Add or Change Local User
Device/Directory Access.
Note: You can also type smit sb_cfglocaccess at the command line.
2. Enter the name of the user whose access you want to add or change.
8-2
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
The default entry is all. Press Enter to select this entry, change the entry to
another user name, or press F4 to list user records previously configured and
select a user from the list.
The following screen shows the system default permission record for all users
on the system:
Add or Change User Device/Directory Access
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
* User name to allow
If all: Users to deny access
Tape or Virtual Devices for backups or "all"
Directory path for backup images
Directory path for network install images
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
F3=Cancel
F7=Edit
Enter=Do
[Entry Fields]
all
[]
[all]
[/usr/lpp/sysback/bf/lo>
[/usr/lpp/sysback/image>
+
+
+
+
F4=List
F8=Image
Figure 8-2. The System Default Permission Record for ″all″ Users
The fields have the following meanings:
User name to allow
The user name entered or selected on the prior screen. You cannot
change this field on this screen.
If all: Users to deny access
Default=(blank/none). If you selected to add or change the record for
“all” users, enter a list of users, separated by spaces, to be explicitly
denied access. In other words, all users except those specified here will
be permitted.
Tape or Virtual Devices for backups or ″all″
Default=(blank/none). If the specified user is allowed to back up to a
tape drive or virtual device, enter the device name or names, separated
by spaces. You can also press F4 to list the available tape and virtual
devices on the system and select one or more devices from the list
using F7.
Enter “all” to provide a backup option for this user for all tape and
virtual devices on the system.
Only the devices entered here are displayed on the SMIT menus when
the user selects a backup.
Directory path for backup images
Default=(blank/none). If you want to allow the specified user to
perform backups to disk image files, enter the directory or directories,
separated by spaces, that the user can back up to. Press F4 to display a
list of backup image directories already selected, or the default
Chapter 8. Local User Access
8-3
directory of /usr/lpp/sysback/bf/local/username. Highlight each entry
you want to save and press F7. Press Enter to continue.
When performing any SysBack backup, you are provided a directory
backup option for each directory specified. The actual file created in the
directory is named according to the information in Chapter 2, “SMIT
Overview”, on page 2-1.
Directory path for network install images
Default=(blank/none). This field applies only to the “root” user,
because only the root user can create a system backup. Enter
information in this field only when configuring the “root” user or “all”
users if root is not specifically defined.
Enter each directory for which you want to save installation images,
each separated by space. Press the F4 key to display a list of installation
image directories already selected, or the default directory of
/usr/lpp/sysback/images/local. Highlight each entry you want to save
and press F7. Press Enter to continue.
When performing the option to Backup the System, the user is
provided a directory backup option for each directory specified. Also,
when this or other systems attempt to perform a network installation,
all backup images in the specified directories that the client system has
permission to read are displayed as network install options.
3. Press Enter to complete this process. Repeat the above steps to grant access to
other users.
Listing User Access
Use the List User Access option to display a list of all user permission records. The
list includes users permitted and denied, along with the devices and/or directories
that each user is permitted to access. To list the users:
1. From the Local User Access menu, select List User Access.
Note: You can also list users by typing smit lb_lslocaccess at a command
line.
The list is displayed with no further prompts.
Removing User Access
Use the Remove User Access option to remove permission records.
Removing a specific user’s permission record does not automatically deny that
user access if there is also a permission record defined for “all” users. If a record
exists for “all” users, and you want to deny access to a specific user, you must
remove the user’s permission record, if any, and then add that user name to the list
of users to deny using the Add or Change Local User Device/Directory Access
option.
To remove user access:
1. From the Local User Access menu, select Remove User Access.
Note: You can also type the fastpath smit sb_ucfglocaccess at a command
line.
2. Select a user from the list of users with permission records currently defined.
8-4
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
If a record exists for “all” users, an option for “all” is displayed. If you remove
the record for “all” users, all users except those with permission records
explicitly defined for them will be denied access.
3. Press Enter to confirm your choice. Doing so removes the selected permission
record from the file with no further prompts.
Chapter 8. Local User Access
8-5
8-6
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Chapter 9. Remote Services
Remote Services is an optional feature of SysBack that enables a host attached to a
network using TCP/IP to read and write backups to and from either tape drives or
disks attached to a remote host. This enables a single host to act as a backup server
for other hosts on the network that do not have a local tape drive attached.
With Remote Services, a backup server machine can enable specific hosts and users
on those hosts to access the local devices or directories. Remote Services also
enables client hosts to be installed from a device or disk image file on a network
installation server.
To use Remote Services, SysBack must be installed on each host that you want to
act as a backup server, as well as each host that backs up its data to the server.
TCP/IP must also be installed on each host to use Remote Services functions.
To
1.
2.
3.
4.
access the menu for configuration and maintenance of a backup server or client:
Type smit at a command line.
Select System Backup & Recovery for AIX.
Select Configuration Options.
Select Remote Services.
You can also use the following fastpath to access this menu: smit sb_remote.
The following screen is displayed:
Remote Services
Move cursor to desired item and press Enter.
Server Options
Client Options
F1=Help
F9=Shell
F2=Refresh
F10=Exit
F3=Cancel
Enter=Do
F8=Image
Figure 9-1. The Remote Services Menu
This first screen contains two options, one for server configuration and one for
client configuration. The following screens show the options for each. Each of these
options is described in the sections below.
© Copyright IBM Corp. 2002
9-1
Server Options
Move cursor to desired item and press Enter.
Configure Remote Services
Add or Change Client Host Access to this Server
List Client Access
Remove Client Access
F1=Help
F9=Shell
F2=Refresh
F10=Exit
F3=Cancel
Enter=Do
F8=Image
Figure 9-2. Server Options from the Remote Services Menu
Client Options
Move cursor to desired item and press Enter.
Configure Remote Services
Add or Change List of Remote Backup Servers
List Remote Backup Servers
Remove a Remote Backup Server
F1=Help
F9=Shell
F2=Refresh
F10=Exit
F3=Cancel
Enter=Do
F8=Image
Figure 9-3. Client Options from the Remote Services Menu
Understanding Remote Host and User Access Permissions
SysBack is very flexible as to which hosts, and users on those hosts, can access the
devices or directories on the local system, as well as which devices or directories
they may access. For instance, specific devices or directories or both can be
assigned to all users on all client hosts, or all devices can be assigned to specific
users on specific hosts.
The following example describes the permission records and how they are
interpreted. Note that this is a complex example. Ordinarily, records are defined for
all hosts, or for each specific host, but not both.
9-2
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Assume the following client host/user records exist:
Host to
allow
Hosts to
deny
Users to
allow
venus
root
venus
all
all
root
all
mars pluto
all
Users to deny
bobby cindy
devices
backup directories
all
/usr/lpp/sysback/bf
venus/root
/dev/rmt1
/usr/lpp/sysback/
images/all
all
greg marsha
peter
network
installation image
directories
/dev/rmt0
vdev0
/usr/lpp/sysback/bf/
all/all
When a client host attempts to gain access to a server, permission records are
checked on the server in the following order:
1. First, the permissions on the server are checked to ensure that the host
attempting to gain access has not been explicitly denied access. Hosts denied
access are shown only under the record assigned to “all” hosts. If, for example,
any user on host mars that attempts to perform a backup to this server will
receive a message that permission is denied.
2. Next, permission records are searched to ensure that the specific user has not
been denied access. Users can be denied access under any record that applies to
“all” users. If on any host user marsha attempts to gain access, she will be
denied access because the host record for “all” hosts explicitly denies access to
her. If user cindy on host venus attempts to gain access, she will be denied
because she is explicitly denied access from venus. User cindy can perform a
backup from any other hosts.
3. Assuming the host or user is not explicitly denied access, records are then
searched for one matching the specific host and user. In the above example, if
the root user on host venus attempts to gain access, this user will be allowed to
backup to all devices and the /usr/lpp/sysback/bf/venus/root directory on the
server.
4. If the above record does not exist, records are searched for one matching “all”
users on the specific host. In the above example, if user jan on host venus
attempts to gain access, she will obtain permission from the record assigned to
all users on venus. She will therefore have access to only device /dev/rmt1 and
no backup image directories.
5. Lastly, if none of the above succeed, records are searched for “all” users on
“all” hosts. If user alice on host neptune attempts to perform a backup to the
server, she will gain access under this record. She will therefore be allowed to
back up to devices /dev/rmt0 and vdev0 and to backup image directory
/usr/lpp/sysback/bf/all/all.
Also note that, by removing the record for host venus and user all, all users on
venus, except root, still have access to the devices and directories defined under the
record for all users on all hosts. Note also that, by removing this record, users bobby
and cindy also have access under all users on all hosts because they are no longer
explicitly denied.
The entry for network installation image directories applies only to either the root
user, or all users if the root user is not explicitly defined. This entry determines the
directory or directories from which the installation image files will be listed during
a network installation operation.
Chapter 9. Remote Services
9-3
Backing Up, Listing, Verifying, or Restoring from Remote Backup
Devices
After Remote Services is configured and remote devices are defined, the only
differences in backing up, listing, verifying or restoring data to or from a remote
tape drive or file is the initial selection of the device itself. When any backup
option is selected from the SMIT menus, a list of available backup devices or
directories is displayed.
After a remote server is configured, any device or directories on the server, for
which the client has been granted access, are listed as well. The only difference is
that the hostname of the server replaces the prefix “Tape,” “VirDev,” or “Dir”
(indicating a tape device, virtual device, and directory respectively).
The SMIT options List Backup Contents, Verify a Backup, or Restore Data from a
Backup display a list of devices and a list of available disk image files (backups to
directory) for all directories on the server the client has access to.
When a server device or directory is selected, the server hostname is transferred to
the Hostname of server field on the options screen. Then, SMIT includes The
hostname of the server in the command to execute.
When executing any SysBack command from the command line, you need to add
the “-h hostname” flag and parameter to the command to reference a remote,
rather than local, device, file or directory.
Configure Remote Services
Before any machine can back up to a remote host, both the client and tape server
machines must have Remote Services configured. When selecting this menu option,
you are prompted to continue. Press Enter to continue configuring Remote
Services.
To configure Remote Services:
1. On the Remote Services menu, select either Server Options or Client Options,
depending on which you are configuring.
2. Select Configure Remote Services.
3. When you are asked if you want to continue, press Enter.
You must perform the above steps for each system that you want to use Remote
Services. These steps are required before you can select any of the following
options.
Adding or Changing Client Host Access to This Server
Before any client host can perform a backup to a remote server, the server must
first enable permission to the client host and to the specific users on the client
perform the backup operations. Providing access to client hosts does not
automatically provide a backup option to the client. After providing access from
the server, you must then use the option Add or Change List of Remote Backup
Servers on the client to provide a remote backup option in SMIT.
If you want to provide access to the server for all hosts on the network, you can
define a single record that applies to all hosts. When doing so, you can exclude or
deny access to specific hosts. You do this by providing a permission record for a
9-4
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
host called “all”. You might not want certain hosts to have access to all devices and
backup directories, so you can deny specific hosts access or change the device or
directory list that specific hosts are allowed to use. The devices and directories
specified when using this option determine the list of backup options that users on
the client host receives when performing backups using SMIT and those devices or
directories that other SysBack commands, executed at the command line, are
permitted to read or write to.
To add or change client host access to the server:
1. From the Server Options menu, select Add or Change Host Access to this
Server.
Note: From the command line, use the fastpath smit sb_cfgremaccess.
2. Enter the host name for the client system. This can be the name of the client
host or “all” to apply the record to all hosts. You can also press F4 to generate
a list of hosts already configured to change an existing record. The default
entry is “all.”
Note: You can enter either the hostname or a full domain name. If you are
operating in a domain network, use the full domain name of the client
host. Using only the host name in a domain network allows access to
any host on the network with the given host name, even though they
might be in different domains.
3. Enter the user name. The default entry is “all”. Press Enter to select this entry,
change the entry to another user name, or press F4 to list user records
previously configured for the selected host and select a user from the list.
The host name and user name selected are transferred to the following options
screen. This example shows the defaults when configuring a permission record
for the root user on host venus:
Add or Change Client Host Access to this Server
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
* Hostname of client system
If all: Hosts to deny access
* User name to allow
If all: Users to deny access
Tape or Virtual Devices for backups or "all"
Directory path for backup images
Directory path for network install images
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
F3=Cancel
F7=Edit
Enter=Do
[Entry Fields]
titanic
[]
all
[]
[]
[]
[]
+
+
+
+
+
F4=List
F8=Image
Figure 9-4. The defaults When Configuring a Permission Record
4. The fields on the above screen have the following meanings. You can change
the fields specified.
Chapter 9. Remote Services
9-5
Hostname of client system
The hostname entered in the prior screen. You cannot change this field
from this screen.
If all: Hosts to deny access
Default=(blank/none). If you selected to add or change the record for
“all” hosts, enter a list of hosts, separated by spaces, that are to be
explicitly denied access. In other words, all hosts except those specified,
will be permitted. When you deny access to a specific host, all users on
that host are also denied access.
User name to allow
The user name entered in the prior screen. You cannot change this field
from this screen.
If all: Users to deny access
If you selected to add or change the record for “all” users, enter a list
of users, separated by spaces, that are to be explicitly denied access. In
other words, all users except those specified will have access.
If you selected to add or change the record for “all” hosts, the users
entered are denied from all hosts. Otherwise, the users entered are
denied only from the selected host.
Tape or Virtual Devices for backups or ″all″
Default=(blank/none). If the specified user is allowed to back up to a
tape drive or virtual device, enter the device name or name, separated
by spaces. Press F4 to list the available tape and virtual devices on the
system and select one or more devices from the list using F7.
Only the devices entered here are displayed on the SMIT menus when
the selected host and user performs a backup to this server.
Directory path for backup images
Default=(blank/none). If you want to allow the specified user to
perform backups to disk image files, enter the directory or directories,
separated by spaces, that the user can back up to. Press F4 to display a
list of backup image directories already selected, or the default
directory of /usr/lpp/sysback/bf/local/username. Highlight each entry
you want to save and press F7. Press Enter to continue.
When performing any SysBack backup, you provided a directory
backup option for each directory specified. The actual file created in the
directory is named according to the information in Chapter 2, “SMIT
Overview”, on page 2-1.
Directory path for network install images
Default=(blank/none). This field applies only to the “root” user,
because only the root user can create a system backup. Enter
information in this field only when configuring the “root” user or “all”
users if root is not specifically defined.
Enter each directory for which you want to save installation images,
each separated by space. Press the F4 key to display a list of installation
image directories already selected, or the default directory of
/usr/lpp/sysback/images/local. Highlight each entry you want to save
and press F7. Press Enter to continue.
When performing the option to Backup the System, you are provided
a directory backup option for each directory specified. Also, when this
or other systems attempt to perform a network installation, all backup
9-6
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
images in the specified directories that the client system has permission
to read are displayed as network install options.
5. Press Enter to complete this process. Repeat the above steps for each host and
user.
Listing Client Access
The List Client Access option displays a list of all host and user permission
records. The list includes hosts permitted and denied and users permitted and
denied for each host. The devices or directories that each host and user are
permitted to access are also displayed.
To list the client host access, select List Client Access from the Server Options
menu. From a command line, you can list host access by typing smit
sb_lsremaccess.
The list is displayed with no further prompts.
Removing Client Access
Use the Remove Client Access option to remove permission records defined in the
Add or Change Client Host Access to this Server option.
Removing a specific host’s permission records does not automatically deny that
host access if there is also a permission record defined for “all” hosts. Likewise, if
you remove a record containing a list of users to allow, those users are not
necessarily denied access if there is also another record for “all” users. If a record
exists for “all” hosts and you want to deny access to a specific host, you must
remove the host permission record, if any, and then add that host name to the list
of hosts to deny using the Add or Change Client Host Access to this Server
option.
To remove a client permission record:
1. From the Server Options menu, select Remove Client Access.
Note: From a command line, type smit sb_ucfgremaccess.
A list of hosts with permission records currently defined is displayed.
2. Select a host from the list. If a record exists for “all” hosts, an option for “all” is
also shown. If you select to remove the record for “all” hosts, all hosts except
those with permission records explicitly defined for them are denied access.
3. After selecting the client host name, a list of users with records defined for that
host is displayed. Select a user name, or “all” from the list to remove the
selected record.
4. Confirm your choice when asked to do so. This removes the selected
permission record from the file with no further prompts.
Adding or Changing the List of Remote Backup Servers
After a backup server has enabled access to a client host using the option Add or
Change Client Host Access to this Server, the client must identify the servers for
which backup options are to be displayed. Only backup options for backup servers
defined using this option are displayed when selecting to perform a backup, list,
verify or restore operation from SMIT.
Chapter 9. Remote Services
9-7
To add a remote backup server on the client host:
1. From the Client Options menu, select Add or Change List of Remote Backup
Servers.
Note: From the command line, type smit sb_cfgremserver.
2. The following screen is displayed:
Add or Change list of Remote Backup Servers
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
* Hostname of server
Show disk image file backup/restore options?
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
F3=Cancel
F7=Edit
Enter=Do
[Entry Fields]
[kong]
no
+
+
F4=List
F8=Image
Figure 9-5. The Add or Change List of Remote Backup Servers Menu
You can change the following fields on this screen:
Hostname of server
Default=(none/blank). Enter the name of the server to be provided a
backup option or press F4 to list the servers previously added and
select a server from the list. You can use either the full domain name or
just the hostname, as long as the hostname can be resolved by itself.
Show disk image file backup/restore options?
Default=no. With this value set to “no”, no disk image file backup
options are displayed for this server. Set this field to “yes” to display
disk file options on the server, but the server must have defined one or
more directories that are accessible by this client. You are provided a
directory backup option for each directory defined on the server.
Even though the server has a directory defined for this host, you might
not want to present a backup option. In this case, set this field to “no”
and no options are displayed.
3. Press Enter to define this server. You receive an appropriate error message if
the server has not yet provided access to this client. When the server is defined,
all future backup, list, verify and restore options will provide one or more
selections for this server, depending on the number of devices or directories
defined for this client.
9-8
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Listing Remote Backup Servers
This option displays a list of all servers currently defined to this client. This list
includes devices and directories accessible to this client, as defined on the server.
To list the remote servers and accessible devices and directories, select List Remote
Backup Servers from the Client Options menu or type smit sb_lsremserver at a
command line.
The list is displayed with no further prompts.
Removing a Remote Backup Server
Use the Remove a Remote Backup Server option to remove servers defined using
the Add or Change List of Remote Backup Servers option.
Removing a server from this list does not deny future access to this client, but only
prevents remote backup options for this server from appearing on the SMIT menus
when you select backup, list, verify, or restore options.
To remove a server from the client:
1. From the Client Options menu, select Remove a Remote Backup Server.
Note: From the command line, type smit sb_ucfgremserver.
2. Select a server from the list of server hostnames currently defined.
3. Confirm your choice when asked to do so. This removes the selected server
from the list.
Remote Commands Access for use with Pull Backups
The Remote Command Access Options are used to configure a backup client. This
enables a server to remotely initiate a backups (also called a “pull backup.”
For example, the server contacts the client to start the backup. In response, the
client initiates the backup and sends data to the server. The server is actually
“pulling” the backup operation from the client.
Using pull backups lets you initiate and control backup operations from a single
location. You may also automate this backup management using the SysBack
Scheduling and Scripting functions described in Chapter 7, “Scheduled Backups
and Scripts”, on page 7-1.
Pull backups require Remote Services Configuration and Remote Command Access
Configuration. The Remote Services Configuration sends data across the network
to the backup server. The Remote Command Access allows the server to initiate
the pull backup.
The server that initiates the pull backup does not need to be the server where data
is sent. For example, you can configure Remote Service between Machine A, the
client, and Machine B, the backup server receiving the backup, and configure
Remote Command Access between Machine A, the client, and Machine C, the
backup initiation server.
To configure Remote Command Access:
1. Select Remote Command Access.
Chapter 9. Remote Services
9-9
Note: From a command line, type smit sb_acces.
2. From the Remote Command Access Menu, select Configure Remote Services.
Note: From a command line, type smit sb_cfgremsvs.
3. Press Enter at the Are you sure? prompt.
4. Select Add or Change Remote Command Access.
Note: From a command line, type smit sb_accessadd.
5. Type the fully qualified path name to the desired backup command. The
SysBack backup commands are located in Appendix A, “Commands”, on
page A-1. Commands section of this manual, or you may determine the
command by pressing F6 when in a backup menu.
Note: All SysBack backup commands are located in /usr/sbin.
6. Enter the fully qualified DNS names of the server system that will initiate the
backup at the “Hostname of Server System(s).”
7. If the hostname specified is “all,” optionally input any host systems to
specifically deny command access at the “If all: Host(s) to deny access” prompt.
8. Press Enter to confirm your selections.
9-10
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Chapter 10. Virtual Devices
A virtual device enables you to logically group multiple physical devices into a
single “virtual” device for easier reference. For example, if you want to use three
tape drives in parallel to perform your backups, you can configure those drives
into a virtual device description. Therefore, each time you want to backup or
restore from those parallel devices, you need to specify only the virtual device
name.
Note that any tape drive can be configured into one or more virtual devices. For
instance, the same tape drive used in a list of sequential virtual devices can also be
used in a list of parallel virtual devices.
Note: When configuring multiple physical devices into a single virtual device, you
can use tape drives of different types. The only limitation is that all drives
must be configured to the same block size. When you use a parallel or
multi-copy virtual device, the speed of the backup is limited to the speed of
the slowest device.
To access the SysBack menus for managing the virtual devices:
1.
2.
3.
4.
At a command line, type smit.
Select System Backup & Recovery for AIX.
Select Configuration Options.
Select Virtual Devices.
You can also access these menus using the fastpath. To do this, type smit
sb_virdev at a command line.
Virtual Devices
Move cursor to desired item and press Enter.
Add a Virtual Device
List Virtual Devices
Remove a Virtual Device
F1=Help
F9=Shell
F2=Refresh
F10=Exit
F3=Cancel
Enter=Do
F8=Image
Figure 10-1. The Virtual Devices Menu
© Copyright IBM Corp. 2002
10-1
Types of Virtual Devices
SysBack uses the following types of virtual devices:
Sequential
A group of devices that are used sequentially, meaning that when
one device reaches its capacity, the backup or restore automatically
continues on the next device in the list. You can have up to eight
sequential devices in the list. When the last device in the list
reaches its capacity, the user is prompted to change the media in
all of the devices before the backup can continue.
Parallel
A group of devices that are used in parallel. In this case, the data is
“striped” across all of the drives in the list, meaning that the first
cluster of data is written to the first drive, the second to the next
drive, and so forth. The writes are optimized to keep as many
drives as possible streaming at all times for best performance.
Although you can back up in parallel to eight drives at a time,
using more than three drives often exceeds the read performance
of the disk drive you are reading from. After you have exceeded
the read performance during a backup, there is no benefit to
having additional parallel devices. In fact, when the writes
out-perform the reads, tape drives might no longer continue
streaming and might enter a “start-stop” mode, decreasing the
write performance even more. Finding the optimal number of
parallel devices to use for your backups is a matter of trial and
error. If performance decreases after adding an additional device to
the parallel virtual device, you have likely exceeded the read
performance of your disk drive or filesystem.
Note: Always keep in mind when using parallel devices that you
must have the same number of devices available in order to
list or restore the files from the backup.
Multi-Copy
A group of devices that each receive an identical copy of the data
being backed up. Unlike parallel devices, this virtual device type
results in identical copies of the same data. This is useful for
creating multiple backups at one time, either for replicating
systems or for both on-site and off-site storage of backup tapes.
Multi-copy backups often take little more time than a single
backup, because only one copy of the data is being read.
Performance has been optimized to enable several devices to be
written to without decreasing performance. Again, finding the
optimal number of devices without severely impacting overall
backup performance is a matter of trial and error.
Note that each copy created using a multi-copy virtual device is
equivalent to a single sequential backup. Therefore, each of the
backups created with the multi-copy virtual device can only be
read independently using a single tape device name or a sequential
virtual device name.
Volume Numbering
When a device reaches the end of the media, the volume number is changed and a
new volume header is written to the next volume. How the volume number is
incremented differs from one virtual device type to another:
Sequential
10-2
The volume number is always incremented by one. When the first
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
device reaches the end of media, the backup continues on the next
device (volume two). When the last device in the list reaches the
end of volume, the user is prompted to change the volume in all of
the drives before the backup can continue. It is necessary to change
volumes in all drives even if only one of the drives will be used
before the backup completes.
Parallel
The volume number of the first drive begins with A1, the second
drive, B1, and the third, C1. When any of the drives reaches end of
volume, the second digit is incremented by one. For instance, if the
second drive (B) reaches the end of volume, the next volume
inserted in that drive will be B2. In this way, it is easy to determine
which tapes belong with which drives and in what order.
Multi-Copy
The volume number in each drive always starts at one, and each is
always incremented by one. Since the same data is written to all
drives, all volume numbers will also be the same. Note, however,
that all drives might not reach end of volume at the same time.
Therefore, the only difference between the backups in each drive
might be the amount of data on each volume.
Autoloaders
You can define any of the virtual device types as a tape library (or “autoloader”).
By setting a virtual device to autoloader/auto-eject, any time a backup tape reaches
end of media, the tape automatically ejects from the tape drive (if the tape drive
supports auto-eject). If the tape drive has an auto-loader device, the autoloader
senses that the tape has been ejected and automatically removes the tape cartridge
and inserts the next available cartridge. When SysBack senses a new cartridge has
been inserted, the backup continues automatically.
Note that the autoloader is defined for use in sequential mode only. Most
autoloaders support sequential insertion of the tape cartridges. SysBack does not
support random insertion of backup cartridges, but ejects the cartridge and waits
for the autoloader to insert the next sequential cartridge.
You can use multiple autoloaders in conjunction with any of the virtual device
types described above. Also, some autoloaders have multiple tape drives. When
you use two drives in the same autoloader, the loader must be able to monitor
both drives simultaneously and change the tapes in both drives as needed.
Adding a Virtual Device
Creating a virtual device is a matter of selecting one or more physical devices and
selecting the type of virtual device to define.
To create a new virtual device:
1. From the Virtual Devices menu, select Add a Virtual Device.
Note: From a command line, type smit sb_mkvirdev.
The following screen is displayed:
Chapter 10. Virtual Devices
10-3
Add a Virtual Device
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
[Entry Fields]
Sequential
[]
no
[/dev/rmt0 /dev/rmt1]
* Write Policy
Description
Auto-loader/Auto-eject?
* Device name(s)
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
F3=Cancel
F7=Edit
Enter=Do
+
+
+
F4=List
F8=Image
Figure 10-2. The Add a Virtual Device Menu
2. You can change the following fields from the screen above:
Write policy
Default=Sequential. Press Tab to select either “Parallel” or “Multi-copy”
write policy. You can also press F4 to list the three options and select
from the list.
Description
Default=Blank. Although this field is left blank at this time, a generic
description is provided if you do not enter a description. The default
description depends on the write policy:
Sequential
Sequential Virtual Device
Parallel
Parallel Virtual Device
Multi-copy
Multi-copy Virtual Device
Enter a description to override the description above. You do not need
to include the physical device name in the description. Do not use
single (’) or double (″) quotation marks in this field.
Autoloader/Auto-eject?
Default=no. Press Tab for “yes” if this device, or these devices, are
autoloaders. If you set this field to autoloader, the user does not receive
prompts to change the volumes.
Device name(s)
Default=Blank. Enter the device names to include in the list (preceded
by /dev), or press F4 to list the available tape drives and select the
drives using F7. If this device is an autoloader, you can enter only one
device name.
If the write-policy is set to Parallel or Multi-copy, you must enter more
than one device name in this field. If the write-policy is sequential,
entering only one device name treats this device no differently than a
single tape drive, unless you also indicate that the device is an
autoloader.
10-4
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
3. Press Enter to create the virtual device.
The new virtual device name is generated and displayed on the screen. Repeat the
above steps for each additional virtual device you want to add.
Listing Virtual Devices
To list all of the virtual devices currently defined, select List Virtual Devices from
the Virtual Devices menu.
This list shows the virtual device name, whether the virtual device is configured as
an autoloader or auto-eject device, the write policy (s=sequential, p=parallel,
c=multi-copy), the virtual device description, and a list of the physical devices that
make up the virtual device.
Removing a Virtual Device
To remove a virtual device:
1. From the Virtual Devices menu, select Remove a Virtual Device.
Note: From a command line, type smit sb_remexclude.
2. Enter the virtual device name to delete or press F4 to list the available virtual
devices and select from the list.
3. Press Enter to delete the specified virtual device.
Chapter 10. Virtual Devices
10-5
10-6
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Chapter 11. Activity Logging
The SysBack log file provides a mechanism to track the completion status of
backup, restore, and verify operations. The log can be used in conjunction with the
progress indicator or file list options, which are used to monitor the ongoing status
of backup, restore, and verify operations.
The sysback log, implemented in a fixed size and in a circular manner, never
grows beyond a specified size. This is called a “head and tail” log. As the log file
grows, the head of the log approaches the tail as if it were a circle. When the head
catches the tail, the log has reached its maximum size and the oldest entries in the
log are purged to make room for new entries. The length of time SysBack log
entries are held is directly related to the maximum size of the log. Increasing the
log size enables the storage of entries for a longer period of time, chronologically.
You can change the default log size and location and also list the log entries.
Accessing the SysBack Logging Menus
To view the SysBack log entries:
1. From the SysBack Main Menu, select Utilities.
2. From the Utilities menu, select SysBack Logging.
Note: At a command line, type smit sb_logging.
Changing the SysBack Logging Attributes (Size and Location)
The SysBack log file (sysback.log) is located, by default, in the /var/adm/ras/
directory. You can change this location, as well as the log size (4096 bytes or 4
Kbytes by default, the minimum size).
You can increase the log size in increments of 4 Kbytes, provided that there is
sufficient space in the log’s destination directory. If you specify to increase the log
size to a number that is not a 4 Kbyte increment, the log size is automatically set to
the next largest multiple of 4 Kbytes. For example, if you specify a size of 4098, the
size will be automatically set to 8192, because 4098 is not a multiple of 4 Kbytes.
To change or show the SysBack logging attributes:
1. From the SysBack Logging menu, select Change/Show SysBack Logging
Attributes.
Note: At a command line, type smit sb_changelog.
2. The following screen is displayed:
© Copyright IBM Corp. 2002
11-1
Change/Show Sysback Attributes
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
Sysback Log Filename
Sysback Log Size
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
[Entry fields]
[/var/adm/ras/sysback.log]
[8192]
F3=Cancel
F7=Edit
Enter=Do
F4=List
F8=Image
Figure 11-1. The Change/Show SysBack Logging Attributes Menu
The fields have the following definitions:
SysBack Log Filename (Required)
Specifies the fully-qualified path for the log location. The default
location is /var/adm/ras/sysback.log.
SysBack Log Size (Required)
Specifies the size, in 4096 bytes (4 Kbytes) increments, of the log. The
default size is 4096.
Displaying the SysBack Log
Use the Display SysBack Log option to view the contents of the SysBack log.
To display the log, select Display SysBack Log from the SysBack Logging menu.
Note: At a command line, type smit sb_displaylog.
The following is an example of a SysBack log:
============ ALOG /var/adm/ras/sysback.log output ===========
Command: /usr/sbin/mkdirback -f/usr/lpp/sysback/bf/local/all -I06261530
-x -p /home/jdavis/sample.txt
Date: Mon Jun 26 15:31:13 CDT 2000
File "/home/jdavis/sample.txt" does not exist or is not readable.
FAILURE: Files or Directory backup failed to complete.
Command: /usr/sbin/mkdirback -f/usr/lpp/sysback/bf/local/all -I06261534 -x
-p /home/jdavis/.profile
Date: Mon Jun26 15:34:48 CDT 2000
Backup up files ’/home/jdavis/.profile’
Start date is Mon Jun 26 15:34:49 CDT 2000
User is root at
Estimated size is 1 MB
SUCCESS: File/Directory backup completed successfully.
=========== ALOG /var/adm/ras/sysback.log output ===========
The format of the log entry is described below:
11-2
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Command
The backup, restore, or verify command syntax executed.
Date
The date and time that the backup, restore, or verify command syntax
executed.
Error Information
The last fatal error that caused an error in the backup, restore, or verify
command.
Status Described as SUCCESS or FAILURE, indicating that the operation
completed or failed.
Chapter 11. Activity Logging
11-3
11-4
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Chapter 12. Exclude Lists
You might want to exclude certain files, directories, filesystems or logical volumes
from a backup to reduce the time needed to make the backup and the amount of
space on the backup media. If, for instance, you have a directory that contains only
temporary files that are recreated each time an application is started, those files
might be excluded from the backup to save time and space. Another example
might be read-only data files used by an application that never change. In the case
of a system failure, it might be easier to reinstall those data files from another
more stationary backup than to include the data on a daily backup.
You can exclude a single file, a directory (and all files beneath that directory), a
filesystem, or a logical volume from all SysBack backups.
Note: If a directory name is excluded, all files below that directory are also
excluded. This includes any files that might be required for the system to
operate properly and might parse into different filesystems. Use discretion
when excluding directories to prevent creating a system backup that cannot
be used as system installation media.
You cannot use the following characters in the exclude list because they have
special meaning to some AIX commands:
$
*
+
?
^
Although some of these characters can be used in a file or directory name in AIX,
those files or directories cannot be properly excluded when entered in the exclude
list. Only the asterisk (*) character can be used in the exclude list, but this character
represents a wildcard, allowing groups of files or directories with similar names to
be excluded.
The following are examples of using wild cards in the exclude list:
/*test*
All files in the system containing the word “test”.
/*old
All files on the system ending with “old”.
/home/b*
All files under /home starting with a “b”.
/home/t*y
All files under /home starting with “t” and ending with “y”.
Note that any entry starting with “/*” includes all directories on the system, not
just the root (/) directory.
To access the SysBack menus for managing the exclude lists:
1. At a command line, type smit.
2. Select ∏.
3. Select Configuration Options.
4. Select Exclude Lists.
© Copyright IBM Corp. 2002
12-1
You can also access these menus using the fastpath. To do this, type smit
sb_exclude at a command line.
Exclude Lists
Move cursor to desired item and press Enter.
Add a File, Directory, or Logical Volume to Exclude List
Display Current Exclude List
Remove Entries from Exclude List
F1=Help
F9=Shell
F2=Refresh
F10=Exit
F3=Cancel
Enter=Do
F8=Image
Figure 12-1. The Exclude Lists Menu
Adding a File, Directory or Logical Volume to an Exclude List
To add a file, directory or logical volume to the exclude list, thereby excluding this
data from all SysBack backups:
1. From the Exclude Lists menu, select Add a File, Directory, or Logical Volume
to Exclude.
Note: From the command line, type smit sb_mkexclude.
The following screen is displayed:
12-2
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Add a File, Directory, or Logical Volume to Exclude List
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
File or directory name to add
OR
Logical Volume name to add
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
[Entry Fields]
[/tmp]
/
[]
/
F3=Cancel
F7=Edit
Enter=Do
F4=List
F8=Image
Figure 12-2. The Add a File, Directory, or Logical Volume to Exclude List Menu
2. Enter data into one of the following fields:
File or directory name to add
Default=(blank/none). Enter a filename or directory name to add to the
exclude list. The file or directory name must begin with a slash (/), but
can contain wildcard (*) characters. If you enter a filename, only that
file is excluded. If you enter a directory name, all files and directories
within the specified directory are excluded.
If you enter a filesystem mount point for the directory, the entire
filesystem is excluded from the backup. This will not, however, prevent
the filesystem from being recreated from backups without restoring the
data.
Logical Volume name to add
Default=(blank/none). Enter the name of a logical volume to exclude
from system and volume group backups. If specified, the logical
volume data is not included in the backup, but you can recreate the
logical volume from the backup without restoring the data.
3. Press Enter to accept the entries. You must repeat the above steps for each file,
directory, or logical volume you want to exclude.
Listing Excluded Files or Directories
To list all of the files, directories or logical volumes currently excluded from
SysBack backups, select Display Current Exclude List from the Exclude Lists
menu. All of the files, directories or logical volumes currently excluded from
SysBack backups are listed.
Chapter 12. Exclude Lists
12-3
Removing Files or Directories from an Exclude List
To remove individual files, directories or logical volumes previously added to the
exclude list:
1. From the Exclude Lists menu, select Remove Entries from Exclude List.
Note: From the command line, type smit sb_rmexclude.
The following screen is displayed:
Remove Entries from Exclude List
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
File or directory name to remove
OR
Logical Volume name to remove
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
[Entry Fields]
[/home/guest/newfiles]
+/
[]
+/
F3=Cancel
F7=Edit
Enter=Do
F4=List
F8=Image
Figure 12-3. The Remove Entries from Exclude List Menu
2. Enter data into one of the following fields:
File or directory name to remove
Default=blank. Enter the full pathname of a file or directory to remove
or press F4 to list all currently excluded files and directories and select
from the list.
Logical Volume name to remove
Default=(blank/none). Enter the name of a logical volume to remove
from the exclude list or press F4 to display a list of previously excluded
logical volumes and select from the list.
3. Press Enter. Any files, directories or logical volumes removed from the exclude
list are now included in all future SysBack backups.
12-4
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Chapter 13. Network Boot/Installation Configuration
The Sysback Network Boot/Install Configuration options create and configure a
host to provide boot and optional installation services to a client host over the
network.
The SysBack Network Boot functions use the BOOTP protocol to enable a client to
communicate with the boot server. The client sends a BOOTP request across the
network to a server. Then, the server, if configured, responds with the information
that the client needs to contact that server and, subsequently, access the network
boot image.
Once the client has successfully transferred the network boot image, the client uses
this image to boot into maintenance mode and display the System Backup &
Recovery for AIX Installation & Maintenance Main Menu.
Classic Network Boot and NIM Resource Boot
This section explains the differences between the Classic Network Boot and the
NIM Resource Boot.
Classic Boot
The SysBack Classic Network Boot method used in versions earlier than 5,
continues in Version 5 for environments that do not use AIX NIM (Network
Installation Manager) resources.
The Classic Network Boot method relies on the / and /usr filesystems of the boot
server for creating the network boot image. The boot server also fulfills other
processing requirements for the client during the boot/installation process.
Therefore, the boot server must meet certain requirements in order for the boot
client to successfully utilize the boot server’s resources. Specifically:
1. The operating system level of the boot server must be the same level as the
boot client when using SysBack only to boot into maintenance mode. (The
client must boot from the same level as what is currently installed.)
2. The operating system level of the boot server must be the same level as the
installation/restore image used to reinstall the client machine when booting
and installing the client. (The client must boot from the same level as the image
to be installed.)
3. The boot server must have installed any AIX filesets required for the client to
support attached hardware.
Note: If filesets are installed specifically to support a boot client, the server
must be rebooted, and the network boot image updated and recreated
before the client will successfully boot from the boot server.
4. The boot server must have installed kernel filesets to support the processor
type (uniprocessor or multiprocessor) of the boot client regardless of whether
or not the boot server is of the same processor type.
Note: If filesets are installed specifically to support a boot client, the server
must be rebooted and the network boot image updated or recreated
before the client can successfully boot from the boot server.
© Copyright IBM Corp. 2002
13-1
5. SysBack must be installed on the boot server.
Note: If filesets are installed specifically to support a boot client, the server
must be rebooted and the network boot image updated or recreated
before the client can successfully boot from the boot server.
NIM Resource Boot
The SysBack NIM Resource Boot method provides a mechanism for the user to use
existing NIM SPOT and LPPSOURCE resources in combination with SysBack
boot/installation operations. The NIM Resource Boot method relies on an existing
NIM SPOT to provide the network boot image and an existing NIM LPPSOURCE
to provide the device support that the client may need to support attached
hardware. Therefore, the following requirements must be met by the NIM
resources:
1. The NIM resources must be correctly configured before attempting to use them
with SysBack. For more information on configuring NIM and its resources,
refer to the book, Network Installation Manager Guide and Reference - Pub#
SC23-4113
2. The NIM SPOT resource server must be the same server to which the client
will send its BOOTP request.
3. The NIM LPPSOURCE resource server can be either the same server, or a
different server, than the NIM SPOT resource server.
4. The boot client has been previously configured as a NIM client.
5. The NIM SPOT resource meets the operating system level, kernel support, and
device support requirements for a boot server using the Classic Network Boot
method described above.
Note: If any filesets are installed to the NIM SPOT resource to specifically
support the boot of a particular client, the network boot image must be
updated or recreated before the client will successfully boot from the
boot server.
6. The NIM LPPSOURCE resource contains any the filesets needed to support
attached hardware, or any kernel support filesets to support processor type
(uniprocessor vs. multiprocessor) that the client will require and are not already
contained within the backup image to be installed.
7. SysBack must be installed in the NIM SPOT resource. For more information on
installing Sysback into a NIM SPOT resource, see “Installing SysBack into a
SPOT Resource” on page 13-10.
Note: If SysBack is installed into a NIM SPOT resource to support SysBack
boot clients, the network boot image must be updated or recreated
before the client can successfully boot from the boot server.
Classic Network Boot
Accessing and Configuring the Classic Network Boot Menus
The Classic Network Boot menus enables the user to:
v
v
v
v
13-2
Add or change a network boot client.
Set network install client defaults.
Rebuild network boot images.
Remove a network boot/install client.
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
To access the Classic Network Boot menus, select Classic Network Boot from the
Network Boot/Install Configuration menu.
Note: From a command line, type sb_netboot.
Configuring or Updating a Network Boot Image
The Network Boot Image exists on the Network Install Server, and is sent to the
client system only upon request in order to boot the client system to the SysBack
Installation and Maintenance Menus. This prevents the need to have boot media,
such as a tape or diskettes, available at the client in order to perform a network
install.
A separate boot image must be created for each type of network adapter to be
used for the network install. The available network install adapter types are
ethernet, token ring, and FDDI. A network boot image must be created for each
platform type as well. For example, an rspc platform can boot from the same
server as an rs6k platform, but each will use a different network boot image.
If you have applied AIX updates to your system, or have added software to
support new hardware on the client, then you should update the boot image using
this option. To update an existing boot image, follow the same instructions as
creating a boot image for the first time. The previous image file will be
overwritten, and any configured clients will begin using the new image
automatically.
Cloning systems: A System Backup created on one machine can be installed on
another machine with a different processor or machine type. If booting a client
machine of a different machine type or with a different hardware configuration,
you must have installed the device support on the server for all devices to be used
by the client machine. You must also have created a separate network boot image
for each platform type.
Adding or Changing a Network Boot Client
This option configures a network boot client that is to be booted from the server.
This option generates the information needed to respond to the client’s BOOTP
request, and assigns the specific boot image that will be used to boot the client,
depending on the client’s network adapter and platform type. If a network boot
image for this client’s platform, kernel, and device type does not already exist, a
boot image will be created in the /usr/lpp/sysback/netinst/bootimages directory.
Important Note: This option provides the client system with permission to mount
the server’s /usr filesystem in read-only mode. If you do not
want the client to be permitted read-only access to the server’s
/usr filesystem once the install has completed, use the Remove a
Network Boot Client option after installation.
If the network adapter changes on the client system, it is necessary to update the
server to reflect the new adapter type and hardware address of the client machine.
You can update the boot client information if already configured, using the steps
for a new client.
The adapter “hardware address” refers to the physical network adapter’s hardware
address, which differs from adapter to adapter. By entering an adapter hardware
Chapter 13. Network Boot/Installation Configuration
13-3
address, the server responds to a broadcast BOOTP request so the client system
will not have to specify server information in the BOOTP menu. This option is not
usually required by most systems.
To add or change a network boot client:
1. From the Classic Network Boot Menu, select Add or Change a Network Book
Client.
Note: From a command line, type smit sb_cfgbootclient.
2. Type the client hostname of the machine that will be booted from this server.
The client hostname must be resolvable by the server.
Note: If changing a client already configured to this server, press F4 to get a
list of previously configured clients.
3. The following screen will be displayed.
Add or Change a Network Boot Client
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
*Client hostname
*Client network adapter type
Client platform/kernel type
Server IP address
Client gateway address (optional)
Client subnet mask (optional)
Client adapter hardware address (optional)
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
F3=Cancel
F7=Edit
Enter=Do
[Entry Fields]
[sysback2.aix.dfw.ibm.c>
token ring
+
rspc
+
[9.19.134.93]
[]
[]
[]
F4=List
F8=Image
Figure 13-1. The Add or Change a Network Boot Client Menu
The options have the following definitions:
Note: If the client selected was previously configured, the fields will retain
their previous values by default.
Client Hostname (Required)
The name of the boot client. This is the hostname, or fully qualified
hostname, of the client machine.
Client Network Adapter Type (Required))
The type of adapter the client uses to boot from the network boot
server. The default is the first network adapter type detected by the
system. The available options are:
v ethernet
v token ring
v FDDI
Client Platform/Kernel Type (Required)
13-4
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
The type of architecture/platform and the processor type of the client
to boot from the network install server. The default is the server’s
platform and kernel type.
The available options are:
chrp
Common Hardware Reference Platform
chrp/MP
Multiprocessor Common Hardware Reference Platform
rs6k
RISC System/6000 (uniprocessor)
rs6k/MP
Multiprocessor RISC System/6000
rspc
PCI-based (PC) RISC System/6000 (uniprocessor)
rspc/MP
Multiprocessor PCI-based (PC) PC RISC System/6000
The first part of the selection is the platform type. The second part
(/MP) indicates the kernel type. If the kernel type portion is omitted, a
uniprocessor (single processor) kernel is used.
Server IP Address (Required)
The IP Address for the boot server. The default is the current IP address
of the system. Change this value only if the boot server has multiple IP
addresses and the default supplied is not desirable.
Client Gateway Address (Optional)
The network address of the gateway that the client must use to reach
this server, if any.
Client Subnet Mask (Optional)
The network mask, if required, that the client must use to communicate
on the network.
Client Adapter Hardware Address (Optional)
The client’s network adapter hardware address. This value is required
for only certain RSPC platform clients or if the user desires to broadcast
the BOOTP request without completing the server information. Do not
use this field for clients that communicate with the server through a
gateway.
4. Once all entries are completed, press Enter to confirm selections and execute
your choice.
Note: If a boot image for the client’s adapter, platform, and kernel combination
does not already exist, a boot image is created automatically when
adding the boot client.
Setting the Network Install Client Defaults
If you install a client system from a network install server, you can customize
installation to reduce, or even eliminate, the amount of information a user must
enter to initiate the installation of the client machine. The complete elimination of
all prompts is referred to as a “no-prompt install” or an “unprompted installation”.
This is accomplished by setting all installation information that the user is
normally prompted for before actually performing the network installation.
Configure the client as a network boot client before you set the network
installation defaults for that client.
Chapter 13. Network Boot/Installation Configuration
13-5
To set the Network Install Client Defaults:
1. From the Classic Network Boot Menu, select Set Network Install Client
Defaults.
Note: From a command line, type smit sb_cfginstclient.
2. Highlight the client and press Enter.
3. The following screen is displayed.
Set Network Install Client Defaults
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
[TOP]
[Entry Fields]
*Client hostname
sysback2.aix.dfw.ibm.c>
Perform no-prompt installation?
no
+
Debug Logging?
no
+
THE FOLLOWING ARE REQUIRED FOR NOPROMPT INSTALL:
Console Device Name
[]
Console Terminal Type (if ascii)
[]
Non-rootvg Volume Group Option
Restore rootvg only +
Restore Device or File
[]
+
Backup Sequence Number (if tape)
[]
#
Remove Sysback from foreign client after install?
no
+
Remove Network Config from foreign client?
no
+
THE FOLLOWING ARE REQUIRED ONLY IF THE INSTALL SERVER DIFFERS FROM THE BOOT
SERVER:
[MORE...11]
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
F3=Cancel
F7=Edit
Enter=Do
F4=List
F8=Image
Figure 13-2. The Set Network Install Client Defaults Menu
The options have the following definitions:
Note: If the client selected was previously configured, the fields retain their
previous values by default.
Client Hostname
This field cannot be changed; it was populated from input of the
previous menu screen.
Perform No-Prompt Installation
Specifies if the client to be installed will be prompted. If this field is set
to “no”, the installation menus will appear after a network boot.
Otherwise, the installation occurs after the network boot without
prompting the user, provided that all required information for the
installation is also provided. The default is “no”.
Debug Logging
Indicates whether or not to turn on debug logging operations for
troubleshooting network boot and installation problems. The default is
“no”.
Console Device Name
Enter the name of the console device prior to installation to prevent
displaying the prompt prior to the installation menus. The console
device can be either an LFT (graphical display) or TTY (ascii display)
13-6
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
device. Examples are /dev/1ft0 or /dev/tty0. Also, if an LFT device exists
on the client, it can also be selected by typing /dev/console. The default
is “none/blank”.
Note: The display must be selected, even for a no-prompt install. This
lets you receive any error or warning messages and prompts you
if the default information provided is insufficient to complete the
installation.
Console Terminal Type (if ascii)
If the console device is an ascii terminal, or TTY device, you must
select a specific terminal type. This ensures that installation menus, if
required, are displayed with the correct terminal emulation. Enter the
device type or press F4 to display a list of all terminal types defined on
the server and select from the list. Examples of common terminal types
are ibm3151, tvi925 and vt100. The default is “none/blank”.
Non-rootvg Volume Group Option
The default is “Restore Rootvg Only”. The following is a description of
the options for this field.
Restore Rootvg Only
Only the rootvg volume group is created and restored from the
media, even if there were other volume groups defined on the
backup. At the end of the install, only the rootvg volume group
will be defined on the system.
Restore All VGs
All volume groups that are included on the backup media are
created and restored. Volume groups that are defined on the
original system but whose data was not included on the
backup are created but data is not restored.
Import Non-Rootvgs
The rootvg volume group is created and restored. All other
volume groups, if any, are imported from disks on the client.
This option assumes that the non-rootvg volume groups
already exist on the client. Use this option when restoring the
rootvg volume group after a system failure where no other
volume groups were affected.
Restore Device or File
The options shown are made available using the Add or Change Client
Host Access to this Server option under the Remote Services menu.
You must select an option from this list to prevent being prompted for
an installation device or file during the installation process. The default
is “none/blank”. For more information on defining these resource, see
Chapter 9, “Remote Services”, on page 9-1.
Backup Sequence Number (if tape)
If the client is installed from tape, and the system backup to be
installed is not the first system backup on the tape, enter the backup
sequence number. This number must be between 1 and the total
number of system backups on the tape media. The default is
“none/blank”.
Remove SysBack from Foreign Client After Install
Indicates whether or not to remove the SysBack program from the
system after installation. Choose “yes” when you have not purchased a
license for the client that will be installed. The default is “no”.
Chapter 13. Network Boot/Installation Configuration
13-7
Remove Network Configuration from Foreign Client
Indicates whether or not to remove network specific information
required to configure the host to the network. Choose “yes” when the
machine installed was cloned from a backup image of a machine that is
still present on the network. The default is “no”.
Install Server IP Address
The address of the installation server. If the boot server is also the
installation server, do not change this option. If the backup media from
which the client is installed exists on a different server, enter the IP
address of the installation server. The default is the boot server’s IP
address.
Gateway Address
The address of the boot server. If the boot server is also the installation
server, do not change this option. If the client is installed from server
other than the network boot server and a different gateway address is
required to reach the installation server, enter that gateway IP address.
The default is the boot server’s gateway address.
Network Device Name
If the client is installed using a network adapter different than it was
booted from, enter the device name of the network adapter here. The
default is “none/blank”.
Subnet Mask
If the client is installed using a network different than it was booted
from, enter the subnet mask, if any, required for the network adapter to
contact the installation server. The default is “none/blank”.
Token-Ring Speed (if applicable)
If the installation server uses a different network adapter, and the
network adapter is token-ring, select the correct token-ring speed. If the
boot server is also the installation server, do not set this option. The
default is “none/blank”.
Ethernet Interface (if applicable)
If the install server uses a different network adapter, and the network
adapter is ethernet, select the correct interface type. If the boot server is
also the installation server, do not set this option. The default is
“none/blank”.
Ethernet Connection Type
If the install server uses a different network adapter, and the network
adapter is ethernet, select the correct connection type. If the boot server
is also the installation server, do not change this option. The default is
“none/blank”.
Post-Install Script
To assign a specific post-install script to run at the end of the
installation process, enter the fully qualified path and the name of the
post-install script file. The file must exist on the installation server and
be readable by the client during installation. Refer to Appendix C,
“Creating Scripts for Customizing the System Backup and Install
Process”, on page C-1 for additional information on post-install scripts.
The default is “none/blank”.
LVM Information File
To use a customized LVM information file during client installation,
enter the fully qualified path and name of the LVM information file in
13-8
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
this field. The LVM information file must contain information that is
compatible with the client. The file must exist on the installation server
and be readable by the client during installation. For additional
information on creating an LVM Information file, see Chapter 15,
“Utilities”, on page 15-1. The default is “none/blank”.
4. Once all entries are complete, press Enter to confirm selections and execute
your choices.
Rebuilding Network Boot Images
If software is updated or installed in the boot source, you must update or rebuild
the boot image. For the Classic Boot method, the source would be the /usr of the
boot server. For the NIM Resource Boot method, the source would be the NIM
SPOT resource.
To build a network boot image:
1. From either the Classic Boot Method or NIM Resource Network Boot Method
menu, select Rebuild Boot Images.
Note: From the command line, type smit sb_updboot image
2. Press F4 to list the image and select an image to update.
3. Press Enter to confirm your choices.
This option can be used to update both Classic method or NIM Resource method
boot images.
Removing a Network Boot/Install Client
This option removes a network boot client and all information configured for
network installation from the boot server. Use this option after the client
installation is complete. It prevents the client system from remotely mounting the
/usr filesystem in read-only mode from the server. If there is no reason to prevent
the client system from mounting /usr from the server, the client configuration may
be retained, enabling the client machine to boot from the server as needed for
re-installation or to perform system maintenance.
To remove a network boot client:
1. From the Classic Network Boot Menu, select Remove a Network Boot Client.
Note: From a command line, type smit sb_ucfgbootclient.
2. Highlight the client and press Enter.
3. Once all entries are complete, press Enter to confirm selections and execute
your choices.
NIM Resource Boot
Accessing and Configuring the NIM Resource Menus
The NIM Resource Boot menus let you:
v Install SysBack into a SPOT resource.
v Query for SysBack installation in a SPOT resource.
v Add or change a network boot client.
v Set network install client defaults.
v Rebuild network boot images.
Chapter 13. Network Boot/Installation Configuration
13-9
v Remove a network boot/install client.
To access the NIM Resource Boot menu, select NIM Resource Boot from the
Network Boot/Install Configuration menu.
Note: From a command line, type smit sb_nimboot.
Installing SysBack into a SPOT Resource
The NIM SPOT resource must have SysBack installed in order for the SysBack
boot and installation process to succeed.
Use the Install SysBack into a Spot Resource for a first time installation of
SysBack into a NIM SPOT resource or to update the level of SysBack installed in
that SPOT resource.
To install SysBack into a NIM SPOT Resource:
1. From the NIM Resource Boot menu, select Install SysBack into a Spot
Resource.
Note: From a command line, type smit sb_nimspotcust.
2. Type the device or directory name where the SysBack install image resides.
3. The following screen is displayed.
Install Sysback into Spot Resource
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
[Entry Fields]
/usr/sys/inst.images
[]
[]
*INPUT device/directory for software
*NIM SPOT Resource Name
*NIM LPP Source Name
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
F3=Cancel
F7=Edit
Enter=Do
+
+
F4=List
F8=Image
Figure 13-3. The Install Sysback into Spot Resource Menu
The options have the following definitions:
INPUT device / directory for software
This field cannot be edited; its contents are specified from a previous
selector screen.
NIM SPOT Resource Name
The NIM object name for the SPOT resource where SysBack will be
installed. Press F4 to get a list of previously configured NIM SPOT
resources.
NIM LPP Source Name
The NIM object name for the LPPSOURCE resource that supplies the
SysBack image. Press F4 to get a list of previously configured NIM
LPPSOURCE resources.
13-10
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Note: The NIM LPPSOURCE resource specified should contain AIX
filesets of the same AIX level as the SPOT resource used to
install the client. For example, if the SPOT resource supplies
boot information for an AIX 4.3.3 system, the corresponding
LPPSOURCE should contain AIX filesets that are also at the AIX
4.3.3 level. This is not a requirement for any optional software
packages installed into the LPPSROURCE; each package can
have its own operating system compatibility requirements. The
level of SysBack contained in the LPPSOURCE can be any level
supported by the AIX level of the specified SPOT resource.
4. Once all entries are complete, press Enter to confirm selections and execute
your choices.
Note: The image is first copied from the specified input device or directory and
then copied to the specified LPPSOURCE before it is installed to the
specified SPOT resource.
Querying for SysBack Installation in SPOT Resource
This option can be used to determine if SysBack is installed in to a NIM SPOT
resource and, if so, at what level.
To query for installation:
1. From the NIM Resource Boot menu, select Query for SysBack Installation in
SPOT Resource.
Note: From the command line, type smit sb_querysb.
2. Press Enter.
Adding or Changing a Network boot Client
This option configures a network boot client to be booted from the server. This
option generates the information necessary to boot the client using the specified
NIM SPOT resource. This option also automatically generates a boot image for the
specified client using the named NIM SPOT resource. SysBack obtains the
information regarding the client’s platform and kernel type from the previously
configured NIM machine object for that client.
To update the client’s boot image after installing or updating the software into the
NIM SPOT resource used by the client, repeat this process to change and update
the client’s boot image.
To add or change a network boot client:
1. From the NIM Resource Boot menu, select Add or Change a Network Boot
Client.
Note: From a command line, type smit sb_nimaddclient.
2. Highlight the client and press Enter.
3. The following screen is displayed.
Chapter 13. Network Boot/Installation Configuration
13-11
Add or Change a Network Boot Client
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
[Entry Fields]
sysback2
[] +
[] +
*NIM Client Name
*NIM SPOT Resource Name
*NIM LPP Source Name
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
F3=Cancel
F7=Edit
Enter=Do
F4=List
F8=Image
Figure 13-4. The Add or Change a Network Boot Client Menu
The options have the following definitions:
NIM Client Name
The NIM machine object name for the boot client. This field cannot be
edited; it was generated from input on the previous selector screen.
NIM SPOT Resource Name
The NIM SPOT name that SysBack will use to create the boot image
for the client named above.
Note: The requirements described in the NIM Resource Boot discussion
in “Classic Network Boot and NIM Resource Boot” on page 13-1
must be followed.
NIM LPP Source Name
The NIM LPPSOURCE resource that SysBack uses to obtain necessary
device support for the boot client when cloning from a backup image
that does not contain the device support needed for this client.
4. Once all entries are complete, press Enter to confirm selections and execute
your choices.
Rebuilding Network Boot Images
If software is updated or installed in the boot source, you must update or rebuild
the boot image. For the Classic Boot method, the source would be the /usr of the
boot server. For the NIM Resource Boot method, the source would be the NIM
SPOT resource.
To rebuild a network boot image:
1. From either the Classic Boot Method or NIM Resource Network Boot Method
menu, select Rebuild Boot Images.
Note: From the command line, type smit sb_updbootimage.
2. Press F4 to list the images and select an image to update.
Note: Although you can select individual boot images to rebuild, all the boot
images associated with a single NIM SPOT, even if you choose only one
image.
3. Press Enter to confirm your choices.
13-12
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Removing a Network Boot/Install Client
This option will remove a network boot client and all information configured for
network installation from the boot server/resources.
This affects only SysBack configurations and does not change any previously
defined NIM resource information.
To remove a network boot/install client:
1. From the NIM Resource Boot menu, select Remove a Network Boot/Install
Client.
Note: From a command line, type smit sb_ucfgbootclient.
2. Highlight the client and press Enter.
3. Press Enter to confirm your choice.
Initiating an SP Node Boot and Install
This menu option enables the user to initiate the network boot sequence for an SP
node because the RS/6000 Scalable POWERParallel systems have special
requirements for installing and cloning.
For a complete explanation on how to access and use this menu option, see
Chapter 18, “Changing the Volume Group, Logical Volume and Filesystem
Attributes”, on page 18-1.
Chapter 13. Network Boot/Installation Configuration
13-13
13-14
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Chapter 14. Tape Drives
This chapter contains information on performing various functions on tape drives.
For virtual devices, some functions can be performed for all physical tape drives
defined for the virtual device. Most of these tape drive options can be performed
from other AIX system administration SMIT menus, but are provided within
SysBack.
To access the tape drive menus:
1. At a command line, type smit.
2. From the SMIT menu, select System Backup & Recovery for AIX.
3. Select Tape Drive Options.
Tape Drives
Move cursor to desired item and press Enter.
List All Defined Tape Drives
Configure a Defined Tape Drive
Unconfigure an Available Tape Drive
Change Tape Drive Characteristics
Unload/Eject Tape from Drive
F1=Help
F9=Shell
F2=Refresh
F10=Exit
F3=Cancel
Enter=Do
F8=Image
Figure 14-1. The Tape Drives menu
Listing tape drives
To list all tape drives defined to the system, whether currently available or not,
select List All Defined Tape Drives from the Tape Drive Options menu.
The list includes all drives defined or available on the system. Only physical tape
drives, and not virtual devices, are displayed. The drives labelled “Defined” have
at one time been detected by the system but are currently unavailable, either
because they are turned off, disconnected, or the Unconfigure an Available Tape
Drive option was used against them. Devices labelled “Available” are available to
use.
© Copyright IBM Corp. 2002
14-1
Configuring a defined tape drive
When a tape drive is either just turned on, reattached, or previously made
unavailable, use this option to make the drive available for use.
To configure a defined tape drive:
1. From the Tape Drive Options menu, select Configure a Defined Tape Drive.
2. From the list of defined tape drives, select the drive you want to configure and
press Enter.
If an error occurs, then the system is unable to detect the drive that it had
previously defined. Make sure the tape drive is turned on and properly connected.
Unconfiguring an Available Tape Drive
You can make a tape drive unavailable for use by the system but still retain its
definition in the system. To make a drive unavailable for use:
1. From the Tape Drive Options menu, select Unconfigure a Defined Tape Drive.
2. From the list of available tape drives, select the drive you want to unconfigure
and press Enter.
Changing Tape Drive Characteristics
The characteristics of a tape drive that you can change differ from one drive type
to another. One characteristic that is common to all drives is the block size. Refer
to your tape drive documentation or the SMIT HELP screen (F1) for information
on additional characteristics that can be changed.
To change the block size of a tape drive:
1. From the Tape Drive Options menu, select Change Tape Drive Characteristics.
2. From the list of available tape drives, select the drive whose characteristics you
want to change.
3. A screen similar to the following is displayed:
14-2
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Change / Show Characteristics of a Tape Drive
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
[Entry Fields]
rmt0
8mm5gb
scsi
5.0 GB 8mm Tape Drive
Available
04-C0-00-5,0
scsi0
5,0
[1024]
yes
no
yes
140
20
yes
Tape Drive
Tape Drive type
Tape Drive interface
Description
Status
Location
Parent adapter
Connection address
BLOCK size (0=variable length)
Use DEVICE BUFFERS during writes
RETURN error on tape change or reset
Use EXTENDED file marks
DENSITY setting #1
DENSITY setting #2
Use data COMPRESSION
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
F3=Cancel
F7=Edit
Enter=Do
+#
+
+
+
+
+
+
F4=List
F8=Image
Figure 14-2. The Change/Show Characteristics of a Tape Drive menu
Depending on the type of tape drive, the options might differ from those
above.
To change the block size, enter a new block size in the BLOCK size (0=variable
length) field or press F4 for a list of valid entries.
Different tape drives might provide multiple block sizes for reading or writing
the data. Use the same block size to read a tape that used to originally write
the tape. It is generally more efficient to write tapes using larger block sizes
(such as 1024, 2048) because this will usually increase the speed at which the
data is written to the tape. It also usually increases the capacity of the tape
because fewer “inter-block gaps” are needed between blocks written.
Notes:
a. Only the root user has permission to change the block size of a tape drive.
b. When you create a system backup, the tape block size is temporarily
changed to 512-bytes, because the installation process that reads that tape is
only capable of reading the tape at 512-byte blocks.
Refer to Appendix D, “Device/System-Specific Information”, on page D-1 for
information on recommended tape drive characteristics.
4. Press Enter to change the tape drive characteristics.
Unloading or Ejecting a Tape from the Drive
You can rewind and eject the tape cartridge from either a local or remote tape
drive or virtual device, assuming the device supports auto-eject. If you select a
virtual device, the tape is ejected from all physical devices defined for the virtual
device. Whether the drive supports auto-eject or not, the drive door must be closed
again (or opened and closed) before the drive can be used again.
To eject the tape from a drive or virtual device:
1. From the Tape Drive Options menu, select Unload/Eject Tape from Drive.
Chapter 14. Tape Drives
14-3
2. From the list of tape drives and virtual devices available to both the local and
remote systems, select the device you want to eject and press Enter.
14-4
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Chapter 15. Utilities
The utilities menu contains options for performing special tasks, usually those used
by advanced users or that are performed only periodically.
To access the utilities menu:
1. At a command line, type smit.
2. From the SMIT menu, select IBM Tivoli Storage Manager for System Backup
and Recovery.
3. Select Utilities.
The Utilities menu is shown below. These options change more frequently than
other SysBack options, because new utilities are introduced into SysBack more
often than other functions. Also, items are removed as they become obsolete or are
integrated with other features of the system.
Utilities
Move cursor to desired item and press Enter.
Display Sysback Product Install History
Print or Display System Configuration
Create a Bootable Tape (without data)
Change Sysback Tape Boot Default
Change Read Permission on a Disk Image File
Create or Edit an LVM Information File
Create a Custom Installation Diskette
Change Sysback Program Defaults
F1=Help
F9=Shell
F2=Refresh
F10=Exit
F3=Cancel
Enter=Do
F8=Image
Figure 15-1. The Utilities Menu
The options available on the Utilities menu are described in the following sections.
Displaying SysBack Product Installation History
You might want to know the installation history of SysBack, particularly when
deciding if product updates are required. Using the Display SysBack Product
Install History option, you can create a list containing the dates and levels of the
initial SysBack installation and any product updates to date. The list contains the
main SysBack product “sysback.rte,” as well as any additional product features
that are installed separately.
© Copyright IBM Corp. 2002
15-1
To display the product history, select Display SysBack Product Install History
from the Utilities menu. You can also create this list using the fastpath smit
sb_showlevel at the command line.
Printing or Displaying System Configuration
Use the Print or Display System Configuration option to document the system
configuration, including all information pertaining to the volume groups, physical
volumes, logical volumes, and filesystems. It provides a report that contains all of
the key information that you might need in case of various system failures.
In most cases, you do not need this information when recovering from system
failures, because the SysBack installation process and the remakevg command
recreates volume groups, logical volumes, and filesystems without requiring the
user to have knowledge of the system configuration. This information, however,
might be useful in planning a migration of filesystems to new hardware or in
reorganizing an existing system to take advantage of disk striping, mirroring, or
other features.
To print or display the system configuration:
1. Select Print or Display System Configuration from the Utilities menu.
2. The following options are displayed:
Enter print queue name ([]=display)
The default value is the current default printer as defined by either the
PRINTER environment variable or the first print queue in the
/etc/qconfig file. Enter the name of the print queue where you want to
send the report or leave blank to display on the screen.
Enter filename to save ([]=display)
Default=blank. Enter the full pathname of the file where you want to
save the report or leave blank to display on the screen.
Note: Both of the above fields must be blank to show the report on the screen.
3. Press Enter to print the report.
The report output includes the following information:
Volume group information
One record for each volume group. Contains all volume group attributes
and physical volume sizes. Example:
Volume Group
-----------rootvg
Auto-on?
-------y
PP Size Quorum? Total
Used
Free
PV(s)
------- ------- -------- -------- -------- -----4 MB
y
250 PPs
213 PPs
37 PPs hdisk0
1002 MB
852 MB
148 MB
Logical Volume Information
One record for each logical volume. Contains most logical volume
attributes and a list of the physical volumes each logical volume resides
on. Example:
Volume Group
Logical Volume
Attributes
-------------- ---------------- -----------------------------------------rootvg
hd8
Type:
jfslog
Copies: 1
Upper: 32
LPs:
1
IntraPV: c
StrpSz: 0
MinLps: 0
InterPV: m
PV(s):
hdisk0
15-2
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Physical Volume Information
One record for each physical volume. Contains the physical volume
location, size and a map of the regions on the disk where each logical
volume resides. Example:
PV Name Volume Group
Location
PVID
MB
PPs
-------- ------------- ------------ ---------------- ----- -----hdisk0
rootvg
00-00-0S-0,0 00000218acd0f607 1002 250
LV Name
PPs
Region
-------------------------------hd8
101
|....X.....|
hd6
51-62
|..X.......|
73-96
|..XX......|
hd4
3
|X.........|
102-103
|....X.....|
hd3
105-109
|....X.....|
hd2
5-27
|XX........|
110-207
|....XXXXX.|
222-242
|.........X|
Filesystem information
All filesystem attributes. Note that the BF Spt and AG Size fields are
displayed only on AIX 4.2 or later systems. Example:
Volume Group
Mount Point
-------------- ---------------rootvg
/
/home
/tmp
/usr
/var
/data/files
FragSz NBPI
------ ---4096
2048
512
4096
4096
4096
4096
4096
512
4096
512
4096
Cmprs BF Spt? AG Size?
----- ------- -------no
n
8
no
y
8
no
n
8
no
n
8
no
n
8
no
y
32
Logical Volume Partition Maps
The specific physical partitions on the physical volumes that are used by
each logical volume. This determines the fragmentation of logical volumes.
Example:
Volume Group
Logical Volume Copy# PV Name
-------------- -------------- ----- --------rootvg
hd8
1
hdisk0
hd6
1
hdisk0
1
hdisk0
hd4
1
hdisk0
1
hdisk0
hd3
1
hdisk0
hd2
1
hdisk0
1
hdisk0
1
hdisk0
PP#s
------101-101
51-62
73-96
3
102-103
105-109
5-27
110-207
222-242
PV Region
-----------|....X.....|
|..X.......|
|..XX......|
|X.........|
|....X.....|
|....X.....|
|XX........|
|....XXXXX.|
|.........X|
Creating a Bootable Tape (Without Data)
Use the Create a Bootable Tape (without data) option to create a bootable tape on
the local system or a remote server. This bootable tape is identical to a system
backup except that it does not contain any backup data.
The primary intention of this option is to enable you to create a tape that can be
used to boot a system to the SysBack Installation and Maintenance Menus without
having to perform a full backup. This tape enables the user to perform system
maintenance or to boot from an alternate media if the boot images on a prior
system backup tape are corrupt or do not contain updated device information.
You can also boot from a tape containing required device support software to
perform a network installation from media on a remote server. In this case, you
Chapter 15. Utilities
15-3
must explicitly indicate that the network support be included on the tape. When
booting from tape to perform a network installation, the server on which the
installation media is present does not need the device support installed to support
the client, but the installation media itself must contain the required device
support.
Cloning systems: A boot tape created on one machine can be used on another
machine with a different processor or platform type. The system backup created on
one machine can also be installed on another machine, even if the device
configuration differs. When cloning systems, the machine on which the boot tape is
created must have installed all of the device and system support software for the
destination platform type, processor type, and other required devices.
To create a bootable tape:
1. From the Utilities menu, select Create a Bootable Tape (without data).
Note: From the command line, type smit sb_mkboottape.
2. On the Device Selector screen, highlight the device you want to use and press
Enter.
3. The following screen is displayed:
Create a Bootable Tape (without data)
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
Device name
Hostname of server
Platform/kernel type for tape boot image
Network install support to include
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
F3=Cancel
F7=Edit
Enter=Do
[Entry Fields]
[/dev/rmt0]
[rspc]
[ent]
/
+
+
F4=List
F8=Image
Figure 15-2. The Create a Bootable Tape (without data) Menu
The fields have the following meanings:
Hostname of server
The server hostname if a server option was selected on the prior device
selector screen. You cannot change this field.
Device name
The device, if a tape drive or virtual device was selected. You cannot
change this field from this screen.
Platform/kernel type for tape boot image
Default=current platform/kernel type. This field is displayed only
when you use a tape drive or virtual device. Change this field only if
you are creating a backup that is to be installed (cloned) onto another
15-4
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
platform type or to a machine running a different kernel. Press F4 to
list the platforms/kernels for which tape boot support is installed on
the system and select from the list.
The following platform/kernel types are currently supported:
chrp
Common Hardware Reference Platform
chrp/MP
Multiprocessor Common Hardware Reference Platform
rs6k
RISC System/6000 (uniprocessor)
rs6k/MP
Multiprocessor RISC System/6000
rspc
PCI-based (PC) RISC System/6000 (uniprocessor)
rspc/MP
Multiprocessor PCI-based (PC) RISC System/6000
Network install support to include
Default=(Blank/None). If, after booting a system from this backup tape,
you want to perform a network installation (rather than installation
from this backup), press F4 to display a list of supported network types
and select a network type from the list. You must select the network
type of the adapter that will be used to install the system from the
network server.
4. Press Enter to begin creation of the bootable tape.
Changing SysBack Tape Boot Defaults
Use the Change SysBack Tape Boot Defaults menu to set tape boot defaults.
Setting tape defaults will let you minimize, possibly eliminate, user prompting
during a tape boot or installation process.
This is similar to the Set Network Install Client Defaults menu for network boot
or installation processes.
To set the Tape Boot Defaults:
1. From the Utilities menu, select Change SysBack Tape Boot Defaults.
Note: From a command line, type smit sb_cfginsttape.
2. The following screen is displayed:
Chapter 15. Utilities
15-5
Change Sysback Tape Boot Defaults
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
Perform no-prompt installation
THE FOLLOWING ARE REQUIRED FOR NOPROMPT INSTALL
Console Device Name
Console Terminal Type (if ascii)
Non-rootvg Volume Group Option
Restore Device (if different than boot device)
Backup Sequence Number (if tape)
Remove Sysback from foreign client after install?
Remove Network Config from foreign client?
Debug Logging?
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
[Entry Fields]
yes
[/dev/1ft]
[]
Restore all VGs
[]
[2]
yes
yes
yes
F3=Cancel
F7=Edit
Enter=Do
+
+
+
+
+
#
+
+
+
F4=List
F8=Image
Figure 15-3. The Change SysBack Tape Boot Defaults Menu
3.
The fields have the following definitions:
Perform No-Prompt Installation?
Perform No-Prompt Installation specifies the client to be installed will
be prompted. If this field is set to no, the installation menus will appear
after a network boot. Otherwise, the install will take place after the
network boot without prompting the user if all the required
information has been provided. The default is no.
Console Device Name
To prevent the prompt shown prior to the installation menus that
determines which console to be used for the installation, you must
enter the name of the console device here. This may be either an LFT
(graphical display) or TTY (ascii display) device. Examples are
/dev/1ft0 or /dev/tty0. If an LFT device exists on the client, it may
also be selected by typing /dev/console. The default is none/blank.
Note: You must select the display, even for a promptless install. You
will then be able to receive any error or warning messages and
allow for prompts if the default information provided is
insufficient to complete the install.
Console Terminal Type (if ascii)
If the console device is an ascii terminal, or tty device, you must select
a specific terminal type. This ensures that installation menus, if
required, are displayed with the correct terminal emulation. You may
enter the device type or press F4 to display a list of all terminal types
defined on the server and select from the list. Examples of common
terminal types are “ibm3151”, “tvi925”, and “vt100.” The default is
none/blank.
Non-rootvg Volume Group Option
The default is set at Restore Rootvg Only. The following is a description
of each option.
15-6
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Restore Rootvg Only
Only the rootvg volume group will be created and restored
from the media, even if there are other volume groups defined
in the backup. At the end of the install only the rootvg will be
defined.
Restore All VGs
All volume groups that are included on the backup media will
be created and restored. Volume groups that were defined on
the original system but whose data was not included in the
backup will be created but no data will be restored.
Import Non-Rootvgs
The rootvg volume group will be created and restored. All
other volume groups will be imported from disks on the client.
This option assumes that non-rootvg volume groups already
exist on the client. You typically use this option when restoring
the rootvg volume group after a system failure and no other
volume groups have been affected.
Restore Device or File
The options which appear are those made available using Add or
Change Client Host Access to this Server under the Remote Services
menu. You must select an option from this list to prevent being
prompted for an install device or file during the installation process.
The default is none/blank. For more information on defining these
resource, see Chapter 9, “Remote Services”, on page 9-1.
Backup Sequence Number (if tape)
If the client is to be installed from tape, and the System Backup to be
installed is not the first System Backup on the tape, enter the Backup
Sequence Number. This number must be between 1 and the total
number of System Backups on the tape media. The default is
none/blank.
Remove Sysback from Foreign Client After Install
This yes or no field indicates whether to remove the SysBack program
from the system after an installation. This is useful when the user has
not purchased a license for the client that will be installed. The default
is no.
Remove Network Congfig from Foreign Client
This yes or no field indicates whether to remove network specific
information required to configure the host to the network. This option
is useful when the machine installed was cloned from a backup image
of a machine that is still present on the network. The default is no.
Debug Logging
This yes or no field indicates whether to turn on debug logging
operations for troubleshooting network boot and installation problems.
The default is no.
Changing the Read Permission on a Disk Image File
When you back up to a disk image file, you can set the read permission of the file
to allow only the original host and/or original user to read from the file. You
might need to change this permission at a later time if, for instance, you need to
use an image file with “original host only” permission as a network installation
image for other systems.
Chapter 15. Utilities
15-7
Using the Change Read Permission on a Disk Image File option, you can change
the owner of the file (host and/or user), the backup description, if any, or the host
or user read permission.
Only the following users have permission to change the read permission of a disk
image file:
v Root user on the system where the file physically resides
v Root user on the system from which the file originated
v User on the system who originally wrote the file
To change the permissions of an existing disk image file:
1. From the Utilities menu, select Change Read Permission on a Disk Image File.
2. Select the disk image file you want to change from the list displayed. This list
displays all the files in the backup or installation image directories assigned to
the current user.
Root users can change files that are not on this list. To do this, select any file
from this list and change it on the Change Read Permission on a Disk Image
File screen that follows.
3. The following screen is displayed:
Change Read Permission on a Disk Image File
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
[Entry Fields]
hambone.austin.ibm.com
[/sysback.images/fatrat> +/
[fatrat]
[root]
all hosts
+
all users
+
[]
Hostname (where file resides)
* File Name
Hostname (owner of file)
User name (owner of file)
Host read permission
User read permission
User description
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
F3=Cancel
F7=Edit
Enter=Do
F4=List
F8=Image
Figure 15-4. The Change Read Permission on a Disk Image File Menu
The fields have the following meanings:
Hostname (where file resides)
The name of the host where the file resides, if a remote file was
selected on the prior selection screen. You cannot change this field on
this screen.
File name
The name of the file selected on the prior screen. You cannot change
this field on this screen.
Hostname (owner of file)
Default=(current owner). The hostname.
15-8
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Note that, if the Host read permission field is set to “this host only”,
changing the owner of the file to a new host prevents the original host
from accessing the file. The root user on the server in which the file
physically resides can always access the file.
User name (owner of file)
Default=(current owner). The user name.
Note that, if the User read permission field is set to “this user only”,
changing the owner of the file to a new user will prevent the original
user from accessing the file. The root user on the host that owns the
file, or the root user on the server in which the file physically resides
will always be able to access the file.
Host read permission
Default=(current host read permission). Press Tab to change the field
from its current value to a new value. The possible values are:
this host only The file can be read only by the host that owns the file.
Either the root user on that host or the user who owns
the file, if different, can read it. The root user on the
server where the file physically resides can also access
the file.
all hosts
The file can be read by any host on the network. You
can still limit access permission by the user name.
User read permission
Default=(current user read permission). Press Tab to change the field
from its current value to a new value. The possible values are:
this user only The file can only be read by the user who owns the file.
The root user on the host that owns the file, or root on
the server on which the file physically resides, can also
access the file. Limit read permission of the file to only
the host that owns the file. If not, the file can be read
by this user from any host on the network.
all users
The file can be read by any user. The access permission
can be limited to the host that owns the file. If not, any
user on the network can read the file.
User description
Default=(current user description, if any). The backup description
specified by the user when the backup was created. This description,
up to 60 characters, can be changed by entering the new description in
this field. Do not include single (’) or double (″) quotes in this field.
4. Press Enter to change the file or press F3 key at any time to exit without saving
any changes.
Creating or Editing an LVM Information File
An LVM information file is created by the mkvginfo command as part of all
SysBack backups that contain LVM (logical volume manager) information. This file
is then used either when reinstalling the system from a backup or recreating
volume groups, logical volumes, or filesystems from a backup.
You might want to create your own LVM information file and tailor it to your
needs. This customized LVM information file can then be placed on the backup by
using the -g flag to either the sysback, mkvgback, mkjfsback or mklvback
Chapter 15. Utilities
15-9
commands. The customized LVM information file can also be written to a diskette
and used to replace the information on an existing system backup during a system
installation. Refer to “Creating a Custom Installation Diskette” on page 15-11 in this
chapter for additional details.
To edit an existing LVM information file or create a new LVM information file from
the information on the current system and edit the new file:
1. From the Utilities menu, select Create or Edit an LVM Information File.
2. Enter the name of an existing file or the name of a new file to create.
3. Select the type of backup for the new file. Choose from the following types:
S
System Backup
P
Power System Backup
V
Volume Group Backup
L
Logical Volume or Filesystem Backup
Select the type of file to create based on the backup the file will be used with.
If you select a type that does not match the actual backup media, errors can
prevent the backup media from being restored properly.
4. If you are creating a new file and you specified the system, power, or volume
group backup type, select the volume group whose data is on the backup. To
select a single volume, highlight the volume group and press Enter. To select
multiple volume groups, highlight each volume group and press F7. When you
have selected all entries, press Enter to continue.
It is important to indicate in this field which volume groups are actually
included on the backup. Not doing so prevents the volume group data from
being restored during a system installation. Regardless of which volume groups
are included in the backup data, any volume group can be recreated from this
backup, either during the system installation process or when recreating
volume groups on an active system.
The LVM file is created or checked for consistency with the system configuration,
and then a screen similar to the following is displayed:
15-10
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
+----------------------------------------------------------+
|
Change Volume Group & Logical Volume Information
|
+----------------------------------------------------------+
Select Physical Volumes for Volume Groups
Change Volume Group Attributes
Select Physical Volumes for Logical Volumes
Change Logical Volume Attributes
Change Filesystem Attributes
+------------------------------------------------------------------------------+
| Select this option to change the physical volumes which are assigned to each |
| volume group.
|
+------------------------------------------------------------------------------+
| Use ARROW or Tab keys to move and ENTER to Select. Press ESC for Main Menu. |
+------------------------------------------------------------------------------+
Figure 15-5. The Change Volume Group & Logical Volume Information Menu
The options for volume groups are not displayed when you select to create or edit
a file for a logical volume or filesystem backup. Also, if this is a power backup, the
option for changing filesystem information does not appear. These menus provide
options for changing virtually all attributes for volume groups, logical volumes,
and filesystems. This process is identical to that for reinstalling a system from a
system backup. Therefore, in both cases, the detailed steps for changing this
information is provided in Chapter 18, “Changing the Volume Group, Logical
Volume and Filesystem Attributes”, on page 18-1.
Creating a Custom Installation Diskette
You can customize the system installation process by using a customized
installation diskette.
To create a customized installation diskette:
1. From the Utilities menu, select Create a Custom Installation Diskette.
Note: From the command line, type smit sb_mkinstdskt.
2. The following screen is displayed:
Chapter 15. Utilities
15-11
Create Custom Installation Diskette
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
[Entry Fields]
[/dev/fd0]
no
[]
[]
[]
[]
Output device name
Format diskettes before copying?
Filename of LVM Information File
Filename of Pre-Installation Script
Filename of Post-rootvg Install Script
Filename of Post-Installation Script
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
F3=Cancel
F7=Edit
Enter=Do
+/
+
+
+
+
F4=List
F8=Image
Figure 15-6. The Create Custom Installation Diskette Menu
The fields have the following meanings:
Output device name
Default=diskette device (/dev/fd0). A diskette drive name. If there is
more than one diskette drive on your system, press F4 to display the
available diskette drives and select a drive from the list.
Format diskette before copying?
Default=no. If the diskette in the drive has not been formatted, press
Tab to change this to “yes.”
Filename of LVM Information File
Enter a filename to include an LVM information file on the customized
installation diskette. Doing so replaces the LVM information contained
on the backup and is typically used to apply many changes to the LVM
information without requiring the user to do so using the installation
menus.
Filename of Pre-installation script
Enter a filename to include a pre-installation script on the customized
installation diskette. If it exists, this file is executed prior to the
installation menus appearing on the screen. This might be used, for
instance, to restore other information from the diskette or to perform
some special system configuration not ordinarily performed by the
installation process.
Filename of Post-rootvg installation script
Enter a filename to include a post-rootvg installation script on the
customized installation diskette. If it exists, this file is executed after the
files for the rootvg volume group are restored.
This is commonly used to execute the sample script
/usr/lpp/sysback/scripts/install.postroot_hwmigrate, provided with
SysBack, that performs an automatic installation of device code from
CD-ROM not already included in the system restored from the backup
media.
15-12
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Filename of Post installation script
Enter a filename to include a post-rootvg installation script on the
customized installation diskette. If it exists, this file will be executed at
the end of the installation process before the system reboots. This can
be used, for instance, to execute the sample script
/usr/lpp/sysback/scripts/install.post_rmnet that removes the network
configuration from the restored system data to prevent any prompts at
the end of the SysBack installation process.
You can find more information on the sample installation scripts provided with
SysBack in Appendix C, “Creating Scripts for Customizing the System Backup and
Install Process”, on page C-1.
After this diskette has been created, insert the diskette during the system boot
process. It is read automatically, and any of the above information files provided
are used during the installation process, overriding the same information on the
backup media if it exists.
Changing SysBack Program Defaults
Use the Change SysBack Program Defaults option to tailor the behavior of certain
functions of SysBack. These include the underlying format of the backups, how
read errors are handled when reading backup media, and attributes used to tune
network performance. This section describes the various parameters that can be
changed and the reasons you might choose to do so.
Normally, you do not need to change the default behavior of SysBack. These
parameters should not be changed except by experienced users who understand
the reasons for doing so, as described below.
To access the menus to change the SysBack program defaults, select Change
SysBack Program Defaults on the Utilities menu. You can also access these menus
by typing smit sb_chgsettings at a command line.
The following screen is displayed. The values shown are the system defaults. If
you changed the attributes previously, those values are displayed.
Chapter 15. Utilities
15-13
Change Sysback Program Defaults
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
[MORE...4]
Filesystem Data Backup Command
ERROR HANDLING SETTINGS
Number of data read errors before
Number of data sync errors before
Maximum number of retries on read
Pad missing data with NULLS after
[Entry Fields]
Backup/Restore
read failure
read failure
errors
sync error?
SYSBACK NETWORK SETTINGS
Data Buffer size
TCP Send Buffer size (in bytes)
TCP Receive Buffer size (in bytes)
Force TCP to always send data immediately?
[BOTTOM]
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
[50]
[50]
[2000]
yes
#
#
#
+
[]
[]
[]
no
F3=Cancel
F7=Edit
Enter=Do
+
#
#
#
+
F4=List
F8=Image
Figure 15-7. The Change SysBack Program Defaults Menu
The following sections describe the options available from the Change SysBack
Program Defaults menu.
Backup Format Settings
This section contains options for changing the format of the backup itself. The
default format is “Backup/Restore”, or the AIX backup and restore command
format.
Although SysBack creates a data format to store information about the backup
contents and system information and to enable multi-copy and striped backups,
the underlying command used to read and write files within the filesystems is the
AIX backup command.
SysBack now enables the user to use either the AIX backup command or the AIX
tar command for backing up and restoring filesystem data. When any SysBack
backup is restored, SysBack automatically detects the backup type and uses the
appropriate corresponding restore command to restore the data.
Note: The backup command is the default and should not be changed unless
another command provides function that the backup command does not.
To change the filesystem backup command, press Tab to scroll through the options.
You can also press F4 to display a list of the values. The default is
“backup/restore.”
The backup command is used because of its flexibility over other commands and
because it backs up data regarding files such as the access control list (ACL) and the
trusted computing base (TCB).
The backup command, however, has changed over time in how it backs up and
restores “sparse” files. A sparse file is a file in which blocks of data have been
written non-sequentially, leaving unallocated blocks in the middle of a file.
Previous levels of the backup command did not preserve this sparseness, and a
15-14
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
sparse file, when restored, expands to include all blocks in the middle of the file,
often causing a filesystem to inadvertently run out of space. The backup command
was later changed to preserve this sparseness, but this is often a problem for
customers experiencing the opposite affect.
If a file is created and all blocks are allocated by writing nulls, or “0s,” throughout
the file, these blocks of data are not preserved during a restore, because the goal is
to preserve sparseness. Because of this, a large file might be restored to a very
small size.
SysBack enables you to select the command you want to use to perform backups.
This can be either the AIX backup and restore commands (which preserve
sparseness) or the tar command (which does not).
Error Handling Settings
This section provides options for controlling the way read errors are handled when
reading from SysBack backup media. Although SysBack itself provides a reliable
backup, the media on which the backup is placed can sometimes become corrupt.
The fields have the following meanings:
Number of data read errors before read failure
Default=50. When SysBack encounters a read error from the media device
driver, by default, it attempts to read up to the number of times specified
in the Maximum number of retries on read errors field. If SysBack is
unable to read the data, a read error is produced, and SysBack either skips
the missing data entirely or pads the missing data with NULL bytes, as
defined by the Pad missing data with NULLs after sync errors field.
This option enables you to specify the maximum number of read errors
that are produced before the backup fails. Specify any number up to 32768
in this field or use a zero (0) to indicate that the reading fails after the first
read error.
Number of data sync errors before read failure
Default=50. The backup is read for each buffer, defined by the buffer size
of the backup. At the beginning of each buffer is a special key that ensures
that the data is being read at the correct point. A “data sync” error occurs
when the key is not encountered or the key has an incorrect sequence
number.
When a sync error occurs, SysBack either skips the missing data altogether
or pad the missing data with NULL bytes, as defined by the Pad missing
data with NULLs after sync errors field.
This field determines the maximum number of sync errors that can occur
before the reading fails. The value of this field can be any number up to
32768. Using a value of zero (0) indicates that the reading should abort
after the first sync error.
Maximum number of retries on read errors
Default=2000. When a read error occurs, by default SysBack attempts to
reread the same buffer of data up to the number of times specified by this
field. The reading fails when a read error occurs and has been retried the
number of times indicated. You can enter a number up to 32768. An entry
of zero (0) indicates that no retries should be attempted.
Chapter 15. Utilities
15-15
Note: Most tape devices, including 8MM tape drives, return an error very
quickly when a read error occurs and allows retries to be attempted
from the same data location. Others, such as DDS 4MM tape drives,
take up to 2 minutes to return from a read error. These tape devices
also do not allow read retries, but still take two minutes to return
from an attempt. Therefore, for these, and similar devices, set this
value to zero (0), because retries are not supported and any attempts
appear to pause the reading indefinitely.
Pad missing data with NULLs after sync errors?
Default=yes. When a data sync error occurs, assuming the reading is set up
to continue, the missing data is padded with NULL bytes by default. This
is to ensure that, although the data has been altered, it remains in the
correct alignment.
Note: It is very important for the data to remain in the correct alignment
when restoring raw logical volume backups. If you do not pad sync
errors with NULL bytes, all of the data following the error is
restored to a different location than expected. Volume group,
filesystem, and file/directory backups use an underlying restore
command that is capable of resynchronizing when there is missing
data in the data stream. Therefore, the value of this field is less
relevant when restoring from these backup types. However, the
restore command sometimes fails when it encounters a large stream
of NULL bytes. In this case, change this value to “no.”
SysBack Network Settings
By tuning certain network parameters, you can increase the performance of
SysBack backups and restores. You can set certain values that affect network
performance during SysBack backups and restores without affecting network
performance of other processes using the same network.
Note: Do not change the network settings using this option unless you are familiar
with network tuning. Also, you must apply the same settings to both the
SysBack server and client hosts.
In most situations, the default settings are adequate. The settings appropriate for
achieving the greatest performance on different systems and networks vary widely,
so no specific guidance can be given here. Refer to Appendix D,
“Device/System-Specific Information”, on page D-1 for any specific
recommendations.
The following fields are displayed in the SysBack Network Settings section:
Data buffer size
Default=32768 (32 Kbytes). The size of the buffer of data that is written to
the network socket in a single write operation.
TCP Send Buffer size (in bytes)
Default=16384 (16 Kbytes). The TCP “send buffer” size, equivalent to the
tcp_sendspace parameter of the AIX no command. If the send buffer size is
greater than 64 Kbytes, the RFC1323 TCP parameter, which is equivalent to
the rfc1323 parameter of the AIX no command, is enabled automatically.
TCP Receive Buffer size (in bytes)
Default=16384 (16 Kbytes). The TCP “receive buffer” size, equivalent to the
tcp_recvspace parameter of the AIX no command. If the receive buffer size
15-16
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
is greater than 64 Kbytes, the RFC1323 TCP parameter, which is equivalent
to the rfc1323 parameter of the AIX no command, is enabled automatically.
Force TCP to always send data immediately?
Select “yes” by pressing Tab if you want to send TCP packets immediately.
Otherwise, a value of “no” indicates that small amount of data should be
collected into single packets before being sent.
When all entries are completed, press Enter to apply the changes.
Note: The changes take affect only during SysBack backups, verifies, and restores,
and have no affect on other system network operations. The changes applied
here do not apply to a SysBack system installation; it is not possible to
query the values set during the installation process.
Enabling Large File Support after Migration to AIX 4.2
Note: This SMIT menu option is not displayed on AIX 4.2 or later systems. If
SysBack is installed onto an AIX 4.2 system, large file support will already
be enabled.
Large file support (support for files larger than 2 GB) is available only on AIX 4.2
or later systems. If SysBack was installed on an AIX 4.1 system, and you later
upgraded AIX to 4.2, then you must inform SysBack that AIX 4.2 is now running
and that large files are now supported.
When you select this option, you are asked to confirm your choice. After pressing
Enter, large file support is enabled, and this menu option is permanently removed
from SMIT.
This process ends with an error message if an attempt is made to run the option
on an AIX 4.1 system. No changes are applied.
Chapter 15. Utilities
15-17
15-18
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Chapter 16. System Installation and Maintenance
A backup created using the System Backup option can be used to completely
reinstall the original system from which it was made or to install other machines
with the same or different hardware configurations. During the installation process,
the following functions are provided:
v The volume group, physical volumes, and logical volumes contained on the
backup are compared with the current system hardware configuration. If there
are inconsistencies, such as missing disks or smaller disks that cannot contain
the prior data, you must change the volume group or logical volume
information to fit the new system configuration.
v A simple menu interface is provided to change any volume group, logical
volume or filesystem size or other attribute. This includes selecting the disks
contained in each volume group or the disks where each logical volume will
reside. You can also exclude certain volume groups, logical volumes, or
filesystems from the installation or move logical volumes and filesystems
between volume groups.
v Optionally, individual volume groups, other than rootvg, can be imported from
disk (if they currently exist), ignored, or recreated and restored. By default, they
are created on the original disks (according to the hardware addresses, if they
exist).
v By default, all logical volumes are recreated at their original sizes and on the
same disks and disk locations, if available. If original disk locations are not
available, the logical volumes are recreated where space permits.
v Logical volumes are, by default, recreated using physical partitions that are
contiguous on the disks, unless you specify that the physical partition maps to
be preserved when the backup was created. You can also select to change the
partition policy for select logical volumes during the installation process.
Allowing the partitions to be created contiguously reduces any prior
fragmentation that might have existed, thereby increasing the I/O performance
to those logical volumes.
v The original device configuration is restored if the same adapters are detected on
the new system.
Recovery Installation
One of the options provided on the Utilities menu in the installation menus is to
perform a no-prompt installation. This option enables you to reinstall the operating
system from a system backup without affecting other filesystems and logical
volumes in the rootvg volume group. This option is not intended for upgrading
the operating system level, but recovering from operating system failures. Refer to
“Performing Recovery Installation” on page 16-12 for added details.
No-prompt Installation
If the installation is to be performed from a network installation server, the client
to be installed can be configured for a promptless installation. Refer to the Set
Network Install Client Defaults option in Chapter 13, “Network Boot/Installation
Configuration”, on page 13-1 for details on this configuration.
© Copyright IBM Corp. 2002
16-1
If the client is configured for a no-prompt installation, the SysBack Installation and
Maintenance menu does not appear as described in the following section, but the
installation proceeds without any input from the user. This occurs if all required
default information has been provided and the LVM information on the backup
media is compatible with the system to be installed. If, for example, the prior
physical volumes are not available or if there is not enough disk space on the
client, errors occur, and the installation process defaults to a prompted installation
as described in the remainder of this section.
If the installation defaults and backup data are compatible with the client system,
the installation proceeds as if you selected the Install the System with Current
Settings option from the Main Menu.
Handling Installation Errors
If an error occurs during installation, it occurs in one of the following ways:
1. If the error is a warning message only, the process displays the message and
proceeds normally.
2. If the error is recoverable, a message details the error, and you are given the
option of either continuing the installation process normally or entering a
system maintenance shell. From the maintenance shell (preceded by a ksh>
prompt), you can manually take whatever steps are needed to recover from the
error and then type exit to continue the installation.
3. During a no-prompt installation, codes are displayed in the system LED panel
indicating both the progress of the installation and if any errors have occurred.
If an error that required user intervention occurs, an LED c48 is displayed, the
detailed error is displayed on the system console, and the process defaults to a
prompted installation. Depending on the point of the error, you are either
prompted as described above or placed in the Installation and Maintenance
menus.
The system LED displays various codes during a tape or network boot process, as
well as various status messages during a no-prompt installation. Refer to
Appendix E, “LED Codes”, on page E-1 for details on the LED codes.
The SysBack Installation and Maintenance Menu
To display the SysBack Installation and Maintenance menu, used to initiate a
SysBack system installation, the machine you want to install must first be booted
from either a SysBack system backup tape or a SysBack network boot server.
Note: The Installation and Maintenance menu is not displayed when you perform
a no-prompt installation after booting from a network boot server, unless an
error has occurred in the installation process due to an incompatibility
between the information on the backup and the system to be installed.
The instructions for booting the system vary for each boot type and also differ
greatly depending on the type of the machine to be booted. Detailed instructions
are provided in Appendix B, “Booting a System for SysBack Installation or
Maintenance”, on page B-1 for booting from either tape or network for all of the
following machine types:
v Common Hardware Reference Platform (uni or multi-processor)
v Microchannel-based RISC System/6000 (uniprocessor)
v Microchannel-based RISC System/6000 SMP (multiprocessor)
16-2
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
v PCI-based (RSPC) RISC System/6000 (uni or multi-processor)
v Scalable POWERparallel (SP) Node (uniprocessor)
v Scalable POWERparallel (SP) SMP Node (multiprocessor)
After you have followed these instructions, the following SysBack Installation and
Maintenance menu is displayed:
+----------------------------------------------------------+
|IBM Tivoli Storage Manager for System Backup and Recovery |
+----------------------------------------------------------+
+--------------------------------+
| Installation & Maintenance |
|
M A I N M E N U
|
+--------------------------------+
Change Installation Device
Change Volume Group & Logical Volume Information
Install the System with Current Settings
Utilities Menu
Reboot the System
+------------------------------------------------------------------------------+
| Installation Device: Tape Drive [rmt0]
|
|
|
+------------------------------------------------------------------------------+
| Use ARROW or Tab keys to move and ENTER to Select.
|
+------------------------------------------------------------------------------+
Figure 16-1. The SysBack Installation and Maintenance Menu
The instructions that follow detail the steps for changing the installation options
and performing the installation.
Changing the Installation Device
The default installation device is set to the device the system was booted from and
displayed on the Main Menu. Select the Change Installation Device option to
change the installation device. After selecting this option, you are presented with a
menu similar to the following:
Chapter 16. System Installation and Maintenance
16-3
+----------------------------------------------------------+
|
Change Installation Device
|
+----------------------------------------------------------+
Device
==> /dev/rmt0
/dev/tok0
/dev/ent0
/dev/ent1
Description
Location
5.0 GB 8mm Tape Drive
04-C0-00-5,0
IBM PCI Tokenring Adapter (14101800 04-05
IBM PCI Ethernet Adapter (22100020) 04-B0
IBM 10/100 Mbps Ethernet PCI Adapte 04-04
Autoloader? No
+------------------------------------------------------------------------------+
|
Select one network device or one or more tape devices.
|
|
The "==>" symbol indicates current choice(s).
|
+------------------------------------------------------------------------------+
| Use ARROW or Tab keys to move, ENTER to select or deselect, ESC when done. |
+------------------------------------------------------------------------------+
Figure 16-2. The Change Installation Device Menu
If you selected a network adapter as the installation device, you can select only
one option. However, you can select more than one tape drive as the installation
device if you are installing from a parallel backup (created on a parallel virtual
device), or if you want to use multiple sequential devices (automatically change to
the next device in the list when prior reaches end of volume). For details on the
use of virtual devices, refer to Chapter 10, “Virtual Devices”, on page 10-1.
Important Note: If you are installing from a local parallel virtual device, meaning
the backup was striped across multiple physical devices, you
must select the same number of devices when performing the
installation. You must also insert the tapes in the drives in the
order they appear on the Installation Device menu.
Changing the network installation server after a network boot: If you booted
from a network boot server but want to install from a device or file on a different
network installation server, select the network adapter used to reach the network
installation server. You will then be able to change the network settings used to
reach the new server.
To select a device, highlight the device and press Enter. If the device is a tape
drive, you can also deselect the device by reentering the line number.
If you select a tape device, the Autoloader option is displayed. Press either n or y
to change this option to either “no” (default) or “yes.” If you change this option to
yes, the system will not prompt you to change volumes but will eject the tape
cartridge and wait for the autoloader to insert a new cartridge before continuing
automatically.
Press Esc to end and return to the Main Menu when you have finished your
selection.
Changing the Network Settings
If you selected a network adapter option, the following screen is displayed, based
on the network adapter type selected:
16-4
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
+----------------------------------------------------------+
|
Change Network Settings
|
+----------------------------------------------------------+
Client IP Address
Server IP Address
Gateway IP Address
Subnet Mask
192.168.1.58
192.168.1.55
255.255.255.0
Ethernet Interface:
==> Standard Ethernet Interface
IEEE 802.3 Ethernet Interface
Ethernet Connection Type:
==> BNC: Coax Cable
DIX: 15-pin D-shell Cable
Network Adapter: ent0 [inactive]
+------------------------------------------------------------------------------+
| Enter the subnet mask, if any, in the format "255.255.255.255". Leave this |
| field blank if no subnet mask is used.
|
+------------------------------------------------------------------------------+
| Use ARROWS or TAB to move or INS/DEL to edit. Press ESC to return.
|
+------------------------------------------------------------------------------+
Figure 16-3. The Change Network Settings Options for a Network Adapter
If you selected an Ethernet adapter, you are presented with options similar to the
following:
+----------------------------------------------------------+
|
Change Network Settings
|
+----------------------------------------------------------+
Client IP Address
Server IP Address
Gateway IP Address
Subnet Mask
192.168.1.58
192.168.1.55
255.255.255.0
Ethernet Interface:
==> Standard Ethernet Interface
IEEE 802.3 Ethernet Interface
Ethernet Connection Type:
==> BNC: Coax Cable
DIX: 15-pin D-shell Cable
Network Adapter: ent0 [inactive]
+------------------------------------------------------------------------------+
| Enter the subnet mask, if any, in the format "255.255.255.255". Leave this |
| field blank if no subnet mask is used.
|
+------------------------------------------------------------------------------+
| Use ARROWS or Tab to move or INS/DEL to edit. Press ESC to return.
|
+------------------------------------------------------------------------------+
Figure 16-4. The Change Network Settings Options for an Ethernet Adapter
If you selected a token ring adapter, you receive options similar to the following:
Chapter 16. System Installation and Maintenance
16-5
+----------------------------------------------------------+
|
Change Network Settings
|
+----------------------------------------------------------+
Client IP Address
Server IP Address
Gateway IP Address
Subnet Mask
192.168.1.58
192.168.1.55
255.255.255.0
Token-ring Speed:
4 Megabits
==> 16 Megabits
Network Adapter: tok0 [inactive]
+------------------------------------------------------------------------------+
| Enter the IP address by which the server refers to this host in the format |
| "111.222.333.444". You need not include leading zeros.
|
+------------------------------------------------------------------------------+
| Use ARROWS or TAB to move or INS/DEL to edit. Press ESC to return.
|
+------------------------------------------------------------------------------+
Figure 16-5. The Change Network Settings Options for a Token Ring Adapter
If you selected an FDDI adapter, you receive options similar to those above, except
that the token ring speed option is not displayed. There are no options specific to
the FDDI adapter.
If you booted from a network boot server, the addresses and subnet mask used to
boot the system are displayed. Otherwise, if the boot media was created on a
system previously network-installed using SysBack, the prior network installation
settings are displayed.
To change the client, server or gateway IP address, or the subnet mask, highlight
the line you want to change and enter the new value.
To change the ethernet interface, ethernet connection type or token ring speed,
highlight the desired selection and press Enter.
When all selections are complete, press Esc to return to the previous menu.
Important note: For token ring networks, be certain to select the correct token ring
speed. Not doing so could cause disruption on the entire network
when the token ring adapter is configured.
Changing Volume Group & Logical Volume Information
The SysBack installation process enables you to fully customize the sizes, locations,
and other attributes for volume groups, logical volumes, and filesystems.
If you are installing from a network tape drive or disk image file, the network
installation server is contacted to obtain a list of tape drives, virtual devices, or
disk image files available to this client. If any options are available, a screen titled
“Select Source for Volume Group Data” is displayed. You must select a single
option from which you want the data for the volume groups to be restored.
After selecting to change the system settings, the system verifies the volume group
information from the backup media against the current system configuration. A
screen similar to the following is displayed:
16-6
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
+----------------------------------------------------------+
|
Verifying LVM Information ..
|
+----------------------------------------------------------+
+------------------------------------------------------------------------------+
| COMPLETE
|
+------------------------------------------------------------------------------+
|
|
| The following errors were found in the volume group records:
|
| ----------------------------------------------------------|
| > Physical Volume ID 0016652284a18698 (previously hdisk0 at
|
| location 04-C0-00-4,0) does not exist on this system.
|
| > There are no physical volumes remaining for the rootvg volume
|
| group. You must assign at least one physical volume to rootvg before
|
| continuing.
|
|
|
|
|
|
|
+------------------------------------------------------------------------------+
+------------------------------------------------------------------------------+
| The above errors must be corrected before the installation may be performed. |
|
|
+------------------------------------------------------------------------------+
|
Press ENTER to continue.
|
+------------------------------------------------------------------------------+
Figure 16-6. An Example Backup Installation Screen
The above screen provides an example in which a backup is being installed on a
different system with physical volumes (disks) that do not match the disk
configuration of the original system from which the backup was taken. If there are
no inconsistencies, this screen shows no errors and no prompts appear.
After the verification process is complete, the following screen is displayed,
providing options for changing volume group, logical volume and filesystem
options:
+----------------------------------------------------------+
|
Change Volume Group & Logical Volume Information
|
+----------------------------------------------------------+
Select Physical Volumes for Volume Groups
Change Volume Group Attributes
Select Physical Volumes for Logical Volumes
Change Logical Volume Attributes
Change Filesystem Attributes
+------------------------------------------------------------------------------+
| Select this option to change the physical volumes which are assigned to each |
| volume group.
|
+------------------------------------------------------------------------------+
| Use ARROW or Tab keys to move and ENTER to Select. Press ESC for Main Menu. |
+------------------------------------------------------------------------------+
Figure 16-7. The Change Volume Group & Logical Volume Information Menu
The option for changing filesystem options is not displayed if the backup media is
a power backup. Since the power backup contains only raw logical volumes, no
filesystem information can be changed.
Chapter 16. System Installation and Maintenance
16-7
The current settings for each volume group are displayed. You can change the
settings for a volume group or for the logical volumes within the volume group by
selecting the line number corresponding to the volume group name.
The remaining steps required to change the volume group and logical volume
information is identical to the process for editing the volume group and logical
volume attributes when recreating a volume group or logical volume on a running
system. These instructions are provided in Chapter 18, “Changing the Volume
Group, Logical Volume and Filesystem Attributes”, on page 18-1.
When you finish your changes, press Esc to return to the Main Menu.
Installing the System with Current Settings
When you have made all selections, such as changing of the installation device or
the volume group and logical volume characteristics, you can continue the
installation by selecting Install the System with Current Settings on the Main
Menu.
If you are installing from a network installation server and did not previously select
to edit the volume group attributes, the server is contacted and a list of available
system backup disk image files and installation devices are displayed. You must
select a disk image file or device from which the volume group data is restored.
The system again compares the current hardware configuration with the volume
group and logical volume attributes from the backup, including any changes you
made from the installation menus, to verify that there is adequate space on the
system to continue the installation. If not, a message indicating the problem is
displayed and you return to the Main Menu.
If no inconsistencies are found in the system configuration, preventing the
installation from proceeding, a confirmation screen similar to the following
example is displayed:
+----------------------------------------------------------+
|
Install the System
|
+----------------------------------------------------------+
+------------------------------------------------------------------------------+
| Verifying space requirements ..
|
+------------------------------------------------------------------------------+
| VG Name
Available MB
Used MB
Free MB
|
| --------------------------------------------|
| rootvg
1996 (249 PPs)
1352 (169 PPs)
644 (80 PPs)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
+------------------------------------------------------------------------------+
+------------------------------------------------------------------------------+
| Continuing the installation at this time will overwrite all disks used by
|
| volume groups you chose to create!
|
+------------------------------------------------------------------------------+
|
Do you wish to continue the installation? (y/n)
|
+------------------------------------------------------------------------------+
Figure 16-8. An Installation Confirmation Screen
Press y to continue with the installation or n to return to the Main Menu.
16-8
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
For systems with system keys, you can turn the system key to the normal position
at any time to enable the system to reboot in normal mode without prompting the
user at the end of the installation.
The installation process performs the following steps:
1. The rootvg volume group is created, as well as all logical volumes and
filesystems in the rootvg volume group.
2. The root (/) and /usr filesystem data is restored.
3. The user postroot-install script is executed if it exists. Refer to Appendix C,
“Creating Scripts for Customizing the System Backup and Install Process”, on
page C-1 for details on the postroot installation script.
4. Root volume group post-installation processing occurs. An example of this
processing is updating the device configuration with the new system
configuration and rebuilding the boot logical volumes.
5. Each additional volume group, and the logical volumes and filesystems, are
created.
6. The filesystem and logical data are restored for all remaining logical volumes
and filesystems for which data is present on the backup media.
7. The user post-installation script is executed if it exists. Refer to Appendix C,
“Creating Scripts for Customizing the System Backup and Install Process”, on
page C-1 for details on the post-installation script.
8. If you are installing on a different machine other than the one used to make
the original backup, you are asked if SysBack should be removed from the
machine. Details for this prompt are described in “Removing System Backup
& Recovery for AIX from a Non-licensed Machine”.
9. If you are installing on a different machine than the original backup was
made from and the original machine was configured on a network, you are
asked if you want to remove the network configuration. Details for this
prompt are described in “Removing the Network Configuration” on
page 16-10.
10. If a system key exists and is in the normal position, the system is rebooted
automatically. Otherwise, you are prompted to turn the key to normal position
and press Enter to reboot. You then see the following message:
Re-applying device configuration from previous install...
The system will now shutdown and reboot in order to activate
the device configuration changes...
The system shuts down and reboots once more. When the installation is complete,
the system should look exactly like the original system from which the backup was
made, with the exception of any changes made during the installation process. If,
however, devices were defined on the original system, and the same physical
devices or adapters do not exist on the system on which you are installing on,
those devices will not be defined to the new system.
Removing System Backup & Recovery for AIX from a
Non-licensed Machine
Because System Backup & Recovery for AIX provides the ability to back up one
machine and use that backup to install another, you replicate the SysBack
programs as well. If you do not own a license of the product for the machine that
you are installing, then you must remove SysBack from the system after the
installation.
Chapter 16. System Installation and Maintenance
16-9
To accommodate this, the installation process will check to see if you have installed
on the same system from which the backup was originally created. If so, SysBack
is retained. If not, the following message will appear at the end of the installation
process, just prior to rebooting:
This system was installed from a tape originating from another machine.
If you do note have a license of System Backup & Recovery for AIX for this
machine, you should select to remove SysBack at this time.
Should SysBack be removed from this system? (yes/no) ==>
If SysBack is not licensed on the current machine, type yes to automatically
remove SysBack from the newly installed system. Typing no retains SysBack from
the original backup.
If at a later time you want to remove System Backup & Recovery for AIX from the
system, you can do so by logging in as “root” and typing the following at the
command line:
/usr/lpp/sysback/rmsysback
This removes all SysBack programs and installation history from the system.
Removing the Network Configuration
If you are installing the machine from a backup that was made on a different
machine and the original machine was configured on a network, you are provided
with the current network settings, including the hostname, IP address and
netmask. You can remove the network configuration from this newly installed
system.
The reason for removing the network configuration at this time is to avoid conflicts
with another machine with the same settings that might be currently active on the
network. Allowing a machine installation with a network definition that conflicts
with another active machine might cause severe network problems.
To remove the network configuration at this time, you will need to enter the
correct settings after the installation is complete using the standard SMIT process
for configuring TCP/IP.
Utilities Menu
The Utilities Menu option provides the following options:
16-10
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
+----------------------------------------------------------+
|
Utilities Menu
|
+----------------------------------------------------------+
Start a Maintenance Shell
Import the rootvg Volume Group
Perform Recovery Installation
Post Install Configuration Options
Set Tape Backup Sequence Number to Install From
Copy a System Dump
+------------------------------------------------------------------------------+
| Select this option to start a maintenance shell (ksh). You will return to
|
| the Main Menu after exiting from the shell.
|
+------------------------------------------------------------------------------+
| Use ARROW/Tab keys to move, ENTER to Select, ESC to return to Main Menu.
|
+------------------------------------------------------------------------------+
Figure 16-9. The Utilities Menu
Each of the options provided on this menu are described below.
Starting a Maintenance Shell
This option places you at a shell prompt and enables you to perform system
maintenance tasks. The following message is displayed:
You have entered a maintenance shell. When finished, type "exit" to
return to the Main Menu...
ksh> _
The maintenance shell is provided for the experienced user, typically to perform
system recovery without reinstalling the system. Details for those tasks is beyond
the scope of SysBack function and is not provided here.
You can return to the Main Menu by typing exit.
Importing the rootvg Volume Group
This option provides the ability to gain access to the operating system data,
provided that a rootvg volume group already exists on one or more disks on the
system.
After selecting this option, you are provided a list of physical volumes that exist
on the system and a list of logical volumes that reside on those disks. You must
select only one physical volume containing the rootvg volume group to import.
Some systems might have more than one rootvg volume group.
It is important that you select a physical volume containing a rootvg volume
group. SysBack cannot determine which disks have a rootvg volume group
because the volume group information does not include the name of the volume
group. However, you should be able to determine the rootvg volume group by
looking at the list of logical volumes contained in the volume group.
Chapter 16. System Installation and Maintenance
16-11
After you have selected a rootvg volume group to import, you are asked:
Mount filesystems after importing?
Type y or n to indicate whether or not the filesystems in the rootvg volume group
should be mounted after the volume group is imported. When performing some
maintenance tasks, such as checking and repairing the root (/) or /usr filesystem,
you would not want to mount the filesystems at this time.
After you make your selection, the rootvg volume group is imported, the root (/)
and /usr filesystems are checked and repaired if necessary, and then all filesystems
are optionally mounted. If you chose to mount the filesystems, you are then asked:
Do you want to rebuild the boot logical volume now?
Type y or n to indicate if the boot logical volume should be rebuilt using the AIX
bosboot command. Doing so can recover from many operating system failures but
causes no harm.
You are then placed in a maintenance shell (preceded by the “ksh>” prompt. You
can perform operating system commands from this prompt. When ready, use
normal operating system procedures for rebooting the system.
Performing Recovery Installation
A recovery installation will restore only the operating system on the client by
performing the following steps:
1.
2.
3.
4.
5.
Import a rootvg volume group.
Check, repair if needed, and mount the root (/) and /usr filesystems.
Restore the root (/) and /usr filesystem data from the backup media.
Rebuild the boot logical volume.
Reboot the system.
This process is not intended for updating the operating system on the client. It is
used to restore the primary operating system filesystems from the client’s backup.
It might be possible to restore the data from a backup taken from another machine,
but note that, because only the root (/) and /usr filesystem data is being restored,
there might be incompatibility with other data in the rootvg volume group.
After you select this option, the LVM information is read from the backup media
and you are provided a list of physical volumes that exist on the system and a list
of logical volumes that reside on those disks. You must select only one physical
volume containing the root (/) and /usr filesystems to be restored. Some systems
might have more than one rootvg volume group.
It is important that you select a physical volume containing a rootvg volume
group. SysBack cannot determine which disks have a rootvg because the volume
group information does not include the name of the volume group. However, you
should be able to determine the rootvg volume group by looking at the list of
logical volumes contained in the volume group.
After you select the rootvg volume group to use, the volume group will be
imported. You are then asked:
16-12
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Do you wish to view/change the logical volume or filesystem info?
Type y or n to indicate if you want to view or change the attributes for the root (/)
and /usr filesystems or logical volumes. This includes such attributes as the size
and disk location of these logical volumes. This provides, for example, an effective
tool for reducing the size of an over-expanded /usr filesystem without having to
reinstall the entire operating system. Refer to Chapter 18, “Changing the Volume
Group, Logical Volume and Filesystem Attributes”, on page 18-1 for additional
details.
You have one final prompt to determine if you want to continue the installation
process. Then the recovery installation begins. If the operating system was restored
from a backup taken from a different machine than the client, you might also be
asked whether or not to remove the SysBack product and the network
configuration. Refer to “Installing the System with Current Settings” on page 16-8
for additional information on these prompts. Upon completion, the system reboots
automatically.
Setting Tape Backup Sequence Number to Install From
If you are installing from tape media, and you have “stacked” system backups
onto the tapes, you might want to install from a system backup other than the first
backup on the tape. If so, you must select the correct backup sequence number
using this option. Although no action is performed on the media at this time,
changing the backup sequence number to anything other than “1” (the first system
backup on the media) causes the media to be forwarded to the correct system
backup before any additional installation processing occurs.
After selecting this option, you are prompted for the backup sequence number to
use. You must enter “1” for the first backup on the media, or any number up to
the number of system backups performed to the media. After making your
selection, you return to the Utilities menu.
Posting Install Configuration Options
The Post Install Configuration Options enables you set certain common user
prompts that display during an installation. Setting the response to these prompts
before installation is a convenient way to minimize the intervention required
during normal installation processing - especially when cloning a backup image
from one machine to another.
There are three possible Post Install Configuration Options.
Remove SysBack Program After Install
This yes or no option indicates whether to remove the Sysback program
from the system after an installation. This is useful when the user has not
purchased a license for the client that will be installed. The default is no.
If you did not set this option prior to installation, you will be prompted for
a response as described in “Removing System Backup & Recovery for AIX
from a Non-licensed Machine” on page 16-9 of this chapter.
This option can also be set prior to installation usingthe Set Network
Install Client Defaults Menu described in “Setting the Network Install
Client Defaults” on page 13-5 in Chapter 13, or “Changing SysBack Tape
Boot Defaults” on page 15-5 in Chapter 15.
Remove Network Settings After Install
This yes or no option indicates whether to remove network specific
Chapter 16. System Installation and Maintenance
16-13
information required to configure the host to the network. This option is
useful when the machine installed was cloned from the backup image of a
machine that is still present on the network. The default is no.
This option can also be set prior to installation usingthe Set Network
Install Client Defaults Menu described in “Setting the Network Install
Client Defaults” on page 13-5 in Chapter 13, or “Changing SysBack Tape
Boot Defaults” on page 15-5 in Chapter 15.
Install Additional Device Support
This option lets you specify a CDROM device name, a tape device name,
or a NIM LPPSOURCE directory name. The NIM LPPSOURCE directory
name provides additional device support to install during installation. This
device support, if available in the specific location, will be installed in
circumstances where the needed device support for this machine is
contained in the boot image, not the backup image to be installed.
If this is a NIM Resource Network Boot, SysBack will automatically pulls
device support from the NIM LPPSOURCE specified when the Network
Boot client was added regardless of whether this option is edited. For more
information on configuring a NIM Resource Network Boot, see Chapter 13,
“Network Boot/Installation Configuration”, on page 13-1.
Copying a System Dump
A system dump is recorded when a hardware failure renders the system inoperable
or when the system user initiates a dump manually. The system dump is stored in
a logical volume designated as a dump device. The system always stops after a
system dump is recorded.
This option copies the information from the dump device to a tape or diskette for
later analysis by experienced AIX system support personnel.
Note: A system dump normally contains much more data than what fits on a
diskette and an error occurs when the diskette is full. However, much of the
data needed for analysis is copied to the diskette before the diskette is full.
Note also that the diskette must be formatted before it can be used in this
procedure.
When you first select this option, information regarding the last recorded system
dump is displayed on the screen, such as the following:
Date:
Tue Nov 11 18:58:55 1997
Dump device:
/dev/hd6
Disk location: 04-C0-00-2,0
Size (bytes):
15339008
Status:
Successful
If no system dump information is available, a message is displayed and no further
options are provided. If system dump information is available, you will receive the
following message:
Do you wish to copy the above system dump to an output device? (y/n)
Type y or n to indicate if you want to proceed. Upon proceeding, you will be
provided a list of tape and diskette devices available on the system. After you
select the device, the dump data is copied to the selected media.
16-14
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Note: This procedure must vary on the rootvg volume group to gain access to the
dump logical volume and other commands used to collect operating system
data. Because the rootvg cannot be varied off, you are placed in a
maintenance shell after the dump copy completes, where you can perform
other operating system commands or reboot the system.
Rebooting the System
An option is provided on the Main Menu to Reboot the System if you choose not
to reinstall the system at this time.
You can reboot in normal mode if the system was previously installed and has not
been overwritten. You can reboot in service mode to restart the installation process.
The process of selecting between normal and service mode differs depending on
the platform type:
1. For rs6k and rs6k/MP machines, you can turn the system key between the
normal and service positions before selecting to reboot. The system boots in the
mode selected.
2. For rspc machines, there is no system key. The system boots in normal mode
(from the first available disk) by default. To boot in service mode, follow the
instructions provided in Appendix B, “Booting a System for SysBack
Installation or Maintenance”, on page B-1.
The system shuts down, the screen clears, and the system reboots.
Chapter 16. System Installation and Maintenance
16-15
16-16
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Chapter 17. RS/6000 Scalable POWERParallel (SP) Boot and
Install Utilities
The RS/6000 SP Systems, unlike the non-SP RS/6000 systems, have special
requirements for installing and cloning. The RS/6000 SP System is comprised of a
Control Workstation (CWS) and one or more nodes. Each node and the CWS are
independent machines that work together using a complex system of network
daemons, configuration files, and hardware. The nodes do not have floppy drives
or LED panels consistent with non-SP RS/6000 systems.
SysBack backup and restore operations on the SP are identical to non-SP RS/6000
systems. When these operations are performed across network interfaces, not to
locally attached tape devices, the setup and configuration requirements do not
differ from the non-SP systems. However, the installation process requires some
additional considerations.
This chapter provides the information necessary to fully exploit SysBack
installations in an SP environment.
Advantages of the SysBack SP Boot and Installation Utilities
In previous versions of SysBack, the SP administrator had to perform several
specific SP installation activities manually. For example, if a node was reinstalled,
the administrator had to set the node to “disk” in the SDR to enable SysBack to
complete the installation process rather than PSSP. The node would have to be
manually customized and the PSSP script run. If the node was cloned from the
image of another node, further actions were required, including updating the
/etc/niminfo file and other configuration files.
In SysBack Version 5, the product works with PSSP to handle these SP system
specific issue during the node’s installation process. SysBack sets the SDR’s
bootp_response, run setup_server, perform node conditioning, and runs the PSSP
script. Reinstallation of a node using the SysBack SP boot and installation utilities
results in a fully recovered node.
The administrator continues to perform the usual activities associated with
preparing an SP node to be installed, as well as, the post-installation application
and network specific activities.
SysBack expects to use the same NIM SPOT and LPPLOURCE that is used when
the node is installed with a mksysb image using the default SP installation process.
An indication that the node has been properly prepared for installation is that the
node can be successfully installed with the SP/NIM designated mksysb image
using the standard PSSP installation process.
For more information on how to work with a nodes NIM SPOT and LPPSOURCE,
see the publications:
v PSSP Command and Technical Reference, Volume 2 - Pub# SA22-7351
v PSSP Administration Guide - Pub# SA22-7348
© Copyright IBM Corp. 2002
17-1
How SysBack works with NIM on the SP System
NIM, Network Installation Manager, is an AIX facility used for the installation and
configuration of software through a network interface. SysBack Version 5 contains
a new network boot/install configuration option that uses NIM Resources to
provide the boot image and the location of device support when an image is
cloned.
SysBack can now use an existing, previously configured NIM SPOT and
LPPSOURCE resource for installation. See Chapter 13, “Network Boot/Installation
Configuration”, on page 13-1 for more information.
The SysBack SP boot/install utilities couple with the NIM Resource Network Boot
features to work with the SP.
Unlike previous versions of SysBack, the boot and installation of a SP node is now
initiated using the Initiate SP Node Boot and Install menu option, or the
spsbboot command. The Classic Network Boot options are no longer
recommended for SP node boot/installation.
For information on the Network Installation Manager (NIM), see the book, Network
Installation Manager Guide and Reference - Pub# SC23-1926.
Special Assumptions
Assume the following of SysBack:
v The SP administrator understands and has executed all SP specific installation
preparations before installing a node using SysBack.
v The SP administrator has properly set up and configured the NIM environment
for the SP complex.
v SysBack will use the same NIM SPOT and LPPSOURCE resources during the
boot/install process as a standard PSSP installation.
v You have installed SysBack into the appropriate NIM SPOT resource. See
Chapter 13, “Network Boot/Installation Configuration”, on page 13-1 for more
information on installing SysBack into the NIM SPOT resource.
v You have added the node to be booted/installed as a SysBack network boot
client before initiating the boot/installation process.
v The SP administrator has installed the necessary device support into the
LPPSOURCE before installation. SysBack now automatically installs necessary
device support not contained in the backup image when cloning that image to a
new SP node through the configured NIM LPPSOURCE.
v You initiate the SP node boot/installation using the SysBack menu option
Initiate SP Node Boot and Install or the sbspboot command.
v The SP administrator can perform and execute any application, SP, or Network
specific post-installation activities. This is especially important after cloning an
image to a new node.
17-2
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Accessing the SP Boot and Install Options
You can access the SP Boot and Install menu one of two ways.
Option 1
1. From the SysBack Main Menu, select Configuration Options.
Note: From a command line, type smit sb_config.
2. Select Network Boot/Install Configuration.
Note: From a command line, type smit sb_netboot_choice.
3. Select Initiate SP Node Boot and Install.
Note: From a command line, type smit sb_spboot.
Option 2
1. From the SysBack Main Menu, select Backup & Recovery Options.
Note: From a command line, type smit sb_opt.
2. Select Initiate SP Node Boot and Install.
Note: From a command line, type smit sb_spboot.
Completing the Initiate SP Node Boot and Install Menu
To initiate the SP node boot and installation:
1. Access the menu using one of the two options explained in the previous
section.
2. At the Select a NIM Client screen, select the desired SP node and press Enter.
3. The following screen is displayed:
Initiate SP Node Boot and Install
Type or select values in entry fields.
Press Enter AFTER making all desired changes.
[Entry Fields]
sysback2
[]
[]
[]
*SP Node Name
NIM SPOT Resource Name
NIM LPP Source Name
XWindow’s Display
F1=Help
F5=Reset
F9=Shell
F2=Refresh
F6=Command
F10=Exit
F3=Cancel
F7=Edit
Enter=Do
+
+
+
F4=List
F8=Image
Figure 17-1. The Initiate SP Node Boot and Install Menu
Chapter 17. RS/6000 Scalable POWERParallel (SP) Boot and Install Utilities
17-3
Unprompted and Prompted Installations
SysBack installations are prompted by default. When booting to the SysBack
Maintenance Menu to begin an installation, you can install from a different
installation device, modify logical volume and volume group attributes, or access
other utilities. See Chapter 16, “System Installation and Maintenance”, on page 16-1
for information on these options.
SysBack also lets you define installation defaults to eliminate most prompts during
installation. This is called an “unprompted install”. See Chapter 15, “Utilities”, on
page 15-1 for more information on setting installation defaults. Also see Chapter 16,
“System Installation and Maintenance”, on page 16-1 for more information on
unprompted installations.
SP installations in a non-SysBack environment are generally unprompted.
Therefore, the SP administrator must set up the installation defaults prior to
initiating an unprompted SysBack installation.
By default, a console window or a s1term window does not automatically open
during an unprompted SysBack installation. However, you can monitor the
progress by opening a s1term using one of two methods:
1. From the SP control workstation, type the command
s1term <frame#><slot#>
where frame# represents the node’s frame and slot# represents the node’s slot.
Note: This opens a read-only terminal. Adding -w to the s1term command
results in a writable terminal.
2. Select Open TTY from the HW Perspectives view of the node using the
perspectives command from the SP Control Workstation.
When you install on one node an image created on a different machine (called
“cloning”), a prompted install is recommended.
Troubleshooting Tips
If problems arise during installation, you can use the following resources:
v For unprompted installs, manually open a s1term (console window) to
determine the progress or failure messages of the installation.
v For boot problems, initiate Manual Node Conditioning of the SP node using the
SP nodecond command. For details on executing this process, see the following
publications or contact your SP System Technical Support Organization:
– PSSP Command and Technical Reference, Volume 1 - Pub# SA22-7351
– PSSP Diagnosis Guide - Pub #GA22-7350
– PSSP Administration Guide - Pub# SA22-7348
v Use the HW Perspective view of the node to monitor the progress of the
boot/installation. This view can be accessed using the SP perspectives command
from the SP Control Workstation.
17-4
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Chapter 18. Changing the Volume Group, Logical Volume and
Filesystem Attributes
This section describes the process for changing the volume group, logical volume,
and filesystem attributes prior to creating (or recreating) a volume group, logical
volume, or filesystem. The same instructions apply to the following SysBack
processes:
1. Installation of a system from a SysBack system backup
2. Recreation of a volume group from a system or volume group backup
3. Creation of a logical volume or filesystem from a system, volume group, logical
volume or filesystem backup
You can recreate a volume group, logical volume, or filesystem from within the
SMIT menus or using the remakevg command at the command line.
When you recreate a volume group, logical volume, or filesystem, the information
is retrieved from the backup media and compared with the current system
configuration to determine if the hardware resources are available to recreate the
volume group, logical volume or filesystem as defined on the media. If the
configuration matches, you can change the attributes. If the configuration does not
match, messages indicate the differences, and you must change the volume group
or logical volume attributes to fit the new system.
When you select to recreate a volume group, logical volume, or filesystem on an
active system or when you select to Change Volume Group & Logical Volume
Information when installing from a system backup, a screen similar to the
following is displayed:
+----------------------------------------------------------+
|
Change Volume Group & Logical Volume Information
|
+----------------------------------------------------------+
Select Physical Volumes for Volume Groups
Change Volume Group Attributes
Select Physical Volumes for Logical Volumes
Change Logical Volume Attributes
Change Filesystem Attributes
+------------------------------------------------------------------------------+
| Select this option to change the physical volumes which are assigned to each |
| volume group.
|
+------------------------------------------------------------------------------+
| Use ARROW or Tab keys to move and ENTER to Select. Press ESC for Main Menu. |
+------------------------------------------------------------------------------+
Figure 18-1. The Change Volume Group & Logical Volume Information Menu
© Copyright IBM Corp. 2002
18-1
Notes:
1. The Change Filesystem Attributes option is not displayed if you are installing
a system from a power backup.
2. The Select Physical Volumes for Volume Groups and Change Volume Group
Attributes options are not displayed if you have chosen only to recreate
specific logical volumes and filesystem from a backup.
Press Esc to return to the Main Menu if you are installing a system or to recreate
volume groups, logical volumes, or filesystems on an active system.
The following sections describe each option, as shown above, in detail.
Selecting Physical Volumes for Volume Groups
If you select the first option on the menu, a screen similar to the following is
displayed:
+----------------------------------------------------------+
|
Select Physical Volumes for Volume Groups
|
+----------------------------------------------------------+
VG Name
rootvg
adsmvg
sysbackvg
Option
Create
Create
Create
Physical Volume(s)
hdisk0
hdisk5 hdisk6 hdisk7
hdisk1 hdisk2
+------------------------------------------------------------------------------+
| Select a volume group from the above list.
|
|
|
+------------------------------------------------------------------------------+
| Use ARROW or Tab keys to move and ENTER to select. Press ESC to cancel.
|
+------------------------------------------------------------------------------+
Figure 18-2. The Select Physical Volumes for Volume Groups Menu
Each volume group and a list of physical volumes, if any, currently assigned to the
volume group are listed. From this screen, select the volume group for which to
change the physical volumes. A screen similar to the following is displayed:
18-2
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
+----------------------------------------------------------+
|
Select Physical Volumes for "rootvg" Volume Group
|
+----------------------------------------------------------+
PVname
hdisk0
hdisk1
hdisk2
hdisk3
hdisk4
hdisk5
hdisk6
hdisk7
hdisk8
Location
Volume Group
04-C0-00-4,0 rootvg
04-02-L
sysbackvg
04-02-L
sysbackvg
04-02-L
(free)
04-02-L
(free)
04-02-L
adsmvg
04-02-L
adsmvg
04-02-L
adsmvg
04-02-L
(free)
Description
16 Bit SCSI
SSA Logical
SSA Logical
SSA Logical
SSA Logical
SSA Logical
SSA Logical
SSA Logical
SSA Logical
Disk
Disk
Disk
Disk
Disk
Disk
Disk
Disk
Disk
Drive
Drive
Drive
Drive
Drive
Drive
Drive
Drive
Drive
Megabytes required: 3760
Currently Selected: 4302
+------------------------------------------------------------------------------+
| Select a PV to add or remove from the "rootvg" volume group. You may NOT
|
| select physical volumes currently assigned to a different VG.
|
+------------------------------------------------------------------------------+
| Use ARROW or Tab keys to move and ENTER to select. Press ESC to cancel.
|
+------------------------------------------------------------------------------+
Figure 18-3. A List of Physical Volumes on a System
This screen contains a list of physical volumes on the system and indicates the
volume group to which the physical volume is currently assigned. If a physical
volume is not currently assigned to a volume group, (free) will be indicated in
the volume group column.
Note that the Megabytes required field indicates the amount of disk space
required to contain the logical volumes defined for this volume group. The
Currently Selected field contains the amount of disk space selected for the volume
group. These fields are updated to reflect any changes to the logical volumes or as
physical volumes are added or deleted from volume groups.
You cannot select a physical volume assigned to a volume group, other than the
volume group with which you are working. You can add or remove a physical
volume from the currently selected volume group by performing one of the
following:
v To add a physical volume to the selected volume group, highlight any physical
volume marked (free) and press Enter. The name of the selected volume group
is added to the Volume Group column, and the new megabytes of disk space
selected are added to the Currently Selected field.
v To remove a physical volume from the selected volume group, highlight a line
containing the selected volume group name and press Enter. The volume group
column then indicates (free) and the Currently Selected field is updated to
reflect the new megabytes of disk space assigned to the volume group.
To move a physical volume assigned to the selected volume group to a different
volume group, remove the physical volume from the current volume group, return
to the prior menu, then select the new volume group to which you want to add
the physical volume.
Press Esc to return to the Change Volume Group & Logical Volume Information
menu.
Chapter 18. Changing the Volume Group, Logical Volume and Filesystem Attributes
18-3
Changing Volume Group Attributes
To change the volume group attributes, select Change Volume Group Attributes
from the menu.
A list of volume groups defined on the media is displayed. Select a volume group
and press Enter. A screen similar to the following example is displayed:
+----------------------------------------------------------+
|
Change "adsmvg" Volume Group Attributes
|
+----------------------------------------------------------+
Attribute
Install Option
Auto Varyon at System Startup?
Physical Partition Size (MB)
Quorum Checking?
Concurrent-capable?
Auto-concurrent Varyon?
Value
Create
yes
8
yes
no
***
+------------------------------------------------------------------------------+
| Type "c" to Create, "i" to Import, or "d" to Delete (ignore).
|
| "Create" will create the volume group and restore data (if on the backup). |
+------------------------------------------------------------------------------+
| Use UP/DOWN ARROW or Tab to highlight option to change. Press ESC to return. |
+------------------------------------------------------------------------------+
Figure 18-4. The Change Volume Group Attributes Menu
The following is a description of each field:
Install Option
Indicates whether the selected volume group should be created, imported
or ignored:
Create Enter “c” to completely recreate the volume group on the specified
disks. If installing the system, all logical volume and filesystem
data are restored if the data was included on the backup media.
Import
Enter “i” to import the volume group from the physical volumes
currently selected for the volume group. If so, the selected physical
volumes must currently contain the volume group information.
Delete Enter “d” if you do not want to create the selected volume group.
If set to this value, the volume group will be ignored and the
physical volumes currently assigned to this volume group can be
assigned to another volume group.
This option is not available for the rootvg volume group.
Auto Varyon at System Startup?
Press “y” or “n” to indicate whether the volume group should be
automatically varied on at system startup. This option cannot be changed
for the rootvg volume group.
18-4
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Physical Partition Size (MB)
Enter a physical partition size in megabytes. Valid options are 2, 4, 8, 16,
32, and 64.
If you change the current value, you will receive the following prompt:
Changing the PP size requires the recalculations of all logical volume sizes. Some
LVs may increase in size when using a larger PP size.
Are you sure you want to change the PP size? (y/n)
To not change the size, press “n”. Otherwise, press “y”. All logical volume
sizes will be recalculated because they are based on the physical partition
(logical partition) size of the volume group.
Changing the partition size might require added space to be used by a
logical volume. For instance, assume the current volume group has a
partition size of 4 MB, and a logical volume within the volume group uses
3 partitions (12 MB). Changing the partition size to 8 MB for the volume
group requires that 2 partitions (16 MB) be used for the same logical
volume to ensure adequate space for the prior data.
Quorum Checking?
Enter “y” or “n” to indicate whether or not quorum checking should be in
effect for this volume group. Quorum checking ensures that there are
always a majority of volume group descriptor areas available for this
volume group.
Concurrent-capable?
Enter “y” or “n” to indicate whether or not this volume group should be
concurrent-capable, allowing the physical volumes in the volume group to
be attached to be shared by more than one host.
Auto-concurrent Varyon?
Enter “y” or “n” to indicate whether or not this volume group should be
varied on in concurrent mode. This option is not available unless you also
selected “y” for the Concurrent-capable prompt above.
When all selections are complete, press Esc to return to the prior menu.
Selecting Physical Volumes for Logical Volumes
SysBack, by default, keeps all logical volumes on the same physical volumes as
they previously resided, provided that the same physical volumes exist on the
current system as those that existed on the original system. If not, the AIX Logical
Volume Manager (LVM) automatically determines a default location for each
logical volume in the volume group based on its size and other attributes selected.
You can, however, select specific physical volumes where you want to place each
logical volume. This is valuable because better I/O performance can be achieved
by placing highly-used logical volumes and filesystems on different physical
volumes.
When a logical volume is striped across multiple physical volumes, achieving the
best I/O performance, you must select the specific physical volumes to use. The
number of physical volumes must be a factor of the number of logical partitions
(size) of the logical volume.
Chapter 18. Changing the Volume Group, Logical Volume and Filesystem Attributes
18-5
To select physical volumes where you want to place a logical volume, select Select
Physical Volumes for Logical Volumes. A screen similar to the following is
displayed, showing a list of all logical volumes defined on the backup media:
+----------------------------------------------------------+
|
Select Physical Volumes for a Logical Volume
|
+----------------------------------------------------------+
LV Name
hd5
loglv02
hd8
loglv00
hd6
sblv
hd4
adsmlv
hd1
lv03
lv00
lv04
VG Name
rootvg
adsmvg
rootvg
sysbackvg
rootvg
rootvg
rootvg
adsmvg
rootvg
sysbackvg
sysbackvg
sysbackvg
LPs Copies StrpSz NumPVs MountPoin
1
1
0
1
1
1
0
1
1
1
0
1
2
1
0
1
40
1
0
1
3
1
0
1
4
1
0
1
/
200
1
0
1
/adsmfs
2
1
0
1
/home
52
1
0
1
/home/sysback
50
1
0
1
/home/sysback/build
21
1
0
1
/netscape
MORE..5
+------------------------------------------------------------------------------+
| The above logical volumes are for volume groups selected to CREATE. Select a |
| logical volume from the above list.
|
+------------------------------------------------------------------------------+
| Use ARROW or Tab keys to move and ENTER to select. Press ESC to cancel.
|
+------------------------------------------------------------------------------+
Figure 18-5. A List of All Logical Volumes Defined on the Backup Media
The above screen shows each logical volume and the number of logical partitions,
number of copies, stripe size and number of physical volumes currently assigned
to the logical volume. Select a logical volume and press Enter. A screen similar to
the following is then displayed, showing a list of physical volumes currently
assigned to the volume group in which the logical volume exists:
+----------------------------------------------------------+
|
Select PVs for "adsmlv" Logical Volume
|
+----------------------------------------------------------+
The following are physical volumes in the "adsmvg" volume group:
PVname
hdisk5
==> hdisk6
hdisk7
Location
04-02-L
04-02-L
04-02-L
Description
SSA Logical Disk Drive
SSA Logical Disk Drive
SSA Logical Disk Drive
+------------------------------------------------------------------------------+
| Select PVs to add to or remove from "adsmlv" logical volume.
|
| The "==>" symbol indicates those currently selected.
|
+------------------------------------------------------------------------------+
| Use ARROW or Tab keys to move, ENTER to select or de-select, ESC to cancel. |
+------------------------------------------------------------------------------+
Figure 18-6. A List of Physical Volumes Assigned to the Volume Group
Select a physical volume and press Enter. To deselect a currently selected physical
volume, select the volume and press Enter.
18-6
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Note: Although you can specify specific physical volumes where a logical volume
is placed, there is no guarantee that the logical volume will be placed only
on those disks unless you create the logical volume using a physical
partition map. Verification ensures that adequate space exists in the volume
group for all of the logical volumes, but no checks are made to ensure there
is adequate space on each disk to satisfy the logical volume selections. If
there is not adequate space on the physical volumes when the logical
volume is created, it is created using the first available space in the volume
group.
When changes are complete, press Esc to return to the previous menu.
Changing Logical Volume Attributes
Use this option to change a wide variety of logical volume characteristics, affecting
its size, physical location, and all other attributes.
A screen containing a list of all logical volumes defined on the media is displayed,
such as the following example:
+----------------------------------------------------------+
|
Change Logical Volume Attributes
|
+----------------------------------------------------------+
LV Name
hd5
loglv02
hd8
loglv00
hd6
sblv
hd4
adsmlv
hd1
lv03
lv00
lv04
VG Name
rootvg
adsmvg
rootvg
sysbackvg
rootvg
rootvg
rootvg
adsmvg
rootvg
sysbackvg
sysbackvg
sysbackvg
Pri
03
04
04
04
10
20
20
20
20
20
20
20
LPs MinLPs Copies MountPoint
1
0
1
1
0
1
1
0
1
2
0
1
40
0
1
3
0
1
4
2
1
/
200 107
1
/adsmfs
2
1
1
/home
52
48
1
/home/sysback
50
48
1
/home/sysback/build
21
14
1
/netscape
MORE..5
+------------------------------------------------------------------------------+
| Select a logical volume from the above list.
|
|
|
+------------------------------------------------------------------------------+
| Use ARROW or Tab keys to move and ENTER to select. Press ESC to cancel.
|
+------------------------------------------------------------------------------+
Figure 18-7. A List of All Logical Volumes Defined on the Media
The list of logical volumes includes the creation priority, number of logical
partitions (size), minimum recommended partitions, number of copies, and
filesystem mount point (if any). If a logical volume had previously been changed
so that it would not be created, the line includes the message “*WILL NOT BE
CREATED*”.
To change the attributes of a logical volume, select a logical volume and press
Enter. A screen such as the following is displayed, indicating the current logical
volume attributes:
Chapter 18. Changing the Volume Group, Logical Volume and Filesystem Attributes
18-7
+----------------------------------------------------------+
|
Change "adsmlv" Logical Volume Attributes
|
+----------------------------------------------------------+
+------------------------------------------+-----------------------+
| Original size: 200 LPS (1600 Mb)
| (*) VG Space (LPs)
|
| Minimum size: 107 LPS (856 Mb)
| Avail Used Free
|
| PVs selected: LV: 1 VG: 3
| 805
201
604
|
+------------------------------------------+-----------------------+
Create this LV? *
yes
Inter-policy
m (minimum)
Logical Volume name adsmlv
Stripe size
0
Volume Group name * adsmvg
Use PP map?
no
Size (in LPs) *
200
MWC?
yes
Copies *
1
Write verify?
no
Priority
20
Bad block relocation? yes
Type
jfs
Relocatable?
yes
Intra-policy
c (center)
Maximum PVs
32
Mount Point: /adsmfs
+------------------------------------------------------------------------------+
| Type "n" if you do not want to create this logical volume. If the LV is not |
| created, no data for the LV or its filesystem will be restored.
|
+------------------------------------------------------------------------------+
| Use UP/DOWN ARROW or Tab to highlight option to change. Press ESC to return. |
+------------------------------------------------------------------------------+
Figure 18-8. The Change Logical Volume Attribute Menu
The box at the upper-left portion of the screen contains information that might
assist in making changes to the logical volume size. The Original size field
contains the original size of the logical volume. This field does not reflect any
changes, so you can always return to this value if you inadvertently change the
size. The Minimum size indicates the recommended minimum size of a filesystem
that is required to contain the filesystem data. The PVs selected field contains the
number of physical volumes currently assigned to the volume group for this
logical volume and to the logical volume itself.
The box at the upper-right portion of the screen contains VG Space information
for the volume group to which the logical volume is assigned. This information is
updated to reflect any changes that are made to the logical volume affecting its
size. Also, if the volume group name is changed, the box is updated to reflect the
values for the new volume group.
The following fields can be changed for the logical volume:
Create this LV?
Indicates whether of not this logical volume (and filesystem) will be
created. If you enter “n” to not create the logical volume, no other fields
can be changed. This logical volume is not created and, if installing a
system, the logical volume or filesystem data is not restored from the
backup media.
If the field currently contains “no,” enter “y” to create the logical volume,
and you can then change any other field.
If you change the value of this field, the VG Space information box at the
top of the screen is updated to reflect the amount of space used in the
volume group containing this logical volume.
Logical Volume name
The name of the selected logical volume. You can change the name of the
current logical volume by entering a new logical volume name in this field.
You can only enter a logical volume that is not currently defined on the
active system or on the backup media.
18-8
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Volume Group name
The volume group in which the logical volume currently exists. To move
the logical volume into a different volume group, enter the new volume
group name. The volume group name entered must be currently active or
must be another volume group recreated from the backup media.
If the volume group name is changed, the VG Space information box at
the top of the screen is updated to reflect the new volume group.
If the selected logical volume contains a filesystem, then you must use a
Journaled Filesystem Log (jfslog) logical volume in the new volume group to
mount the filesystem. If the logical volume is moved to a new volume
group, the /etc/filesystems file is updated automatically and the filesystem
is mounted using the first available jfslog logical volume in the new
volume group. The volume group name cannot be changed for a filesystem
logical volume if no jfslog logical volume exists in the new volume group.
Size (in LPs)
The size of the logical volume in logical partitions. The actual disk space
required by this logical volume is the number of logical partitions
multiplied by the number of copies (indicated in the next field). To change
the size of the logical volume, enter a new value.
You cannot change this field if you are installing the system from a power
backup.
If the logical volume contains a filesystem, the filesystem is created at the
same size as the logical volume.
Entering a size smaller than the minimum size displayed at the top of the
screen might cause you to run out of filesystem space when restoring the
data from the backup media. If you are installing a system, over-reducing
the root (/) or /usr filesystems usually results in a system installation
failure.
When you change the size, the VG Space box at the top of the screen is
updated to reflect the new amount of space used in the volume group
containing this logical volume.
Copies
The number of copies (mirrors) of the logical volume. A “1” indicates there
is only one copy or that the logical volume is not mirrored. Changing the
number of copies to greater than one causes the logical volume to be
“mirrored” to another disk. Mirroring is usually implemented to keep a
copy of the logical volume on separate physical volumes to protect data
against a hardware failure.
The number of copies cannot be changed if the Stripe Size is a value other
than 0.
If the number of copies exceeds the number of physical volumes assigned
to the logical volume (or volume group if no physical volumes are
specifically assigned to this logical volume), you will be asked “Do you
wish to allow more than 1 copy on a single PV? (y/n)”. If you answer
“y”, the new number of copies is accepted. If “n”, the new value is
ignored.
When you change the number of copies, the VG Space box at the top of
the screen is updated to reflect the new amount of space used in the
volume group containing this logical volume.
Chapter 18. Changing the Volume Group, Logical Volume and Filesystem Attributes
18-9
Priority
The default priority in which the logical volume will be created. This is
especially important if different logical volumes might be contending for
the same region of a particular disk. Setting the priority of a logical
volume to a lower number gives it a higher priority.
By default, the priority is set for logical volumes based on their types. The
logical volumes appear in the logical volume list in the order in which you
want them created. To create one logical volume before another logical
volume, change the priority to less than the other. The next time the logical
volumes are listed, the changed logical volume appears in a different place
in the list.
Note that this field is a two-digit value, sorted first by the first character,
and then by the second. An entry of “1” is sorted after an entry of “06”, so
use two digits for all entries.
Type
The current logical volume type. The type should not be changed for
paging, jfslog, dump device and boot logical volumes. Although the type “jfs”
is normally used to indicate filesystem logical volumes, this and other
logical volumes, except those specified above, can be changed to any
desired value indicating the type of data contained in the logical volume.
Outer
edge middle center middle
e
m
c
im
spindle
Intra-policy
The current intra-physical volume policy, a policy that determines the
region of the physical volumes where the logical volume should be placed.
The following diagram indicates the available intra-policies:
Inner
edge
ie
Inner
edge
ie
middle
im
Outer
center middle edge
c
m
e
This diagram represents a physical volume (or disk drive) and the
available regions. In the center of the diagram is the spindle of the disk. To
either side of the spindle are the “inner-edge” regions; outside of those are
the “inner-middle” regions. You can enter one the following:
e
edge (or outer-edge)
m
middle (or outer-middle)
c
center
im
inner-middle
ie
inner-edge
Mount Point
The directory that is the mount point for logical volumes containing
filesystems. If the logical volume does not contain a filesystem a dash (-) is
shown in this field and the field cannot be changed.
You cannot change this field if you are installing the system from a power
backup.
To change the mount point of the filesystem, enter a new directory here.
The directory you specified will be created (if it does not already exist)
when the filesystem is created. If installing the system, the data for the
filesystem is restored to the new mount point. For example, if the old
mount point was /home/files and the new mount point is to be /data/files,
then a file previously called /home/files/userinfo is restored as
/data/files/userinfo.
18-10
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
If changed, the /etc/filesystems file is updated to reflect the new mount
point for the filesystem so that all future attempts to mount the filesystem
will do so to the new mount point.
Inter-policy
The current inter-physical volume policy, which indicates whether or not
the logical volume should be spread across multiple disks. This value can
have a different affect depending on the following:
v If no physical volumes are selected for the logical volume: a maximum
value spreads the logical volume across either all of the disks in the
volume group or the value in the Maximum PVs field, whichever is less.
A minimum value uses the first disk in the volume group unless it fills,
in which case the next disk in the volume group is used for the
remainder.
v If you select physical volumes for this logical volume, a maximum value
spreads the logical volume across all of the disks selected for the logical
volume, not to exceed the value of the Maximum PVs field. A minimum
value uses the first disk in the list should be used unless it fills, in which
case the next disk in the list is used.
v If only one disk has been selected for the logical volume or volume
group, this policy has no affect.
v If the Stripe size is set to any value other than zero (0), this policy has
no effect.
Stripe Size
The current stripe size (in K-bytes) or “0” if the logical volume is not
striped. Striping a logical volume means that data is inter-dispersed
throughout the specific physical volumes in blocks of data indicated by the
stripe size. Striping usually benefits random I/O performance, but having
a logical volume spread over more than on physical volume causes the
entire logical volume to be unreadable if any of the disks containing a part
of the logical volume becomes unavailable.
For a logical volume to be striped, you must first assign two or more
physical volumes to this logical volume, and the number of logical
partitions (LPs) must be a multiple of the number of disks selected. You
must enter the stripe size in K-bytes (1024 byte blocks). Valid values are 0,
4, 16, 32, 64 or 128. Setting the value to “0” indicates the logical volume
will not be striped.
You cannot change this field if you are installing the system from a power
backup, if the logical volume is to be created using a partition map (Use
PP map? field), or if the number of Copies is greater than 1.
Use PP map?
Indicates whether or not to preserve physical partition mapping. This
value is set to “no” by default, or to “yes” if you selected to preserve
physical partition mapping when the backup was created.
Select “y” to retain the exact prior physical partition placement of the
logical volume as it was placed on the original system. Creating a logical
volume using the partition map is typically performed only if the logical
volume was originally created using a partition map. If the logical volume
was not originally created using a map, doing so now often retain the
partition fragmentation that exists after logical volumes are periodically
increased in size. This would usually negatively impact I/O performance
to this logical volume.
Chapter 18. Changing the Volume Group, Logical Volume and Filesystem Attributes
18-11
When you use a map to create a logical volume, the prior physical
volumes must still exist and be of the same size. You cannot change the
size, number of copies, striping, or physical volumes for a mapped logical
volume. Also, the Intra-policy and Inter-policy are ignored.
Select “n” to not retain the prior physical partition map. In this case, the
partitions in the logical volume are created contiguously, resulting, in most
cases, in increased I/O performance. The logical volume will, however,
continue to be created in the general location as defined by the physical
volume list, Intra-policy and Inter-policy.
MWC?
Indicates whether or not the mirror-write-consistency checking is in effect. If
so, the volume group status area is updated each time a write is performed
to a mirrored logical partition. This status is then used, in the case of a
disk or system failure, to ensure that all mirrored partitions are identical
when the volume group is again varied on. Turning off the MWC increases
I/O performance when writing to mirrored logical volumes, but requires
you to resynch all mirrors in their entirety in case of an unexpected system
halt to ensure all data between the mirrors is consistent.
Enter “y” or “n” to change the value. This field has no affect for single
copy (unmirrored) logical volumes.
Write-verify?
Indicates whether a read is performed after every write to ensure the data
was written properly. Turning on this attribute decreases the I/O
performance when writing to a logical volume but provides higher
assurance that data was written properly and is readable.
Enter “y” or “n” to change the value.
Bad block relocation?
Indicates whether the software should relocate data in bad blocks when a
write error occurs. This action is normally taken by the disk hardware, but
the hardware might not have this feature or might not have enough space
for the number of blocks to relocate. Turning on this value does not affect
I/O performance.
To change this value, enter “y” or “n”.
Relocatable?
Indicates whether or not this logical volume can be moved to a new
location when a volume group is reorganized using the AIX reorgvg or
migratepv command. If you do not want to enable this logical volume to
relocate, enter “y” in this field.
Maximum PVs
The maximum number of physical volumes onto which the logical volume
can be placed. This value is useful in limiting the number of physical
volumes used when the Inter-policy is set to maximum.
To change the current value, enter a number between 1 and 32.
Changing Filesystem Attributes
Selecting this option enables changes to the filesystem attributes for logical
volumes to contain filesystem data. This option does not appear when you install
from a power backup, because all filesystem data was backed up as raw logical
volumes, and changes to the filesystem attributes are overwritten during the
restoration of the data.
18-12
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
When you first select the option, the following screen is displayed:
+----------------------------------------------------------+
|
Change Filesystem Attributes
|
+----------------------------------------------------------+
LV Name
hd4
adsmlv
hd1
lv03
lv00
lv04
hd3
hd2
instlv
lv01
hd9var
Mount Point
/
/adsmfs
/home
/home/sysback
/home/sysback/build
/netscape
/tmp
/usr
/usr/sys/inst.images
/usr1
/var
+------------------------------------------------------------------------------+
| Select a filesystem from the above list.
|
|
|
+------------------------------------------------------------------------------+
| Use ARROW or Tab keys to move and ENTER to select. Press ESC to cancel.
|
+------------------------------------------------------------------------------+
Figure 18-9. The Change Filesystem Attributes Menu
Select a filesystem to change and press Enter. A screen similar to the following is
displayed, containing the current attributes for the filesystem:
+----------------------------------------------------------+
|
Change Filesytem Attributes for "adsmlv" LV
|
+----------------------------------------------------------+
Attribute
Mount Point
Fragment Size
Number of Bytes per Inode (NBPI)
Use Data Compression?
Large File Support?
Allocation Group (AG) Size
Value
/adsmfs
4096
4096
no
n
8
+
+------------------------------------------------------------------------------+
| Type the new top directory where this filesystem is to be mounted.
|
|
|
+------------------------------------------------------------------------------+
| Use UP/DOWN ARROW or Tab to highlight option to change. Press ESC to return. |
+------------------------------------------------------------------------------+
Figure 18-10. The Current Attributes for a Filesystem
The following attributes can be changed for the filesystem:
Mount Point
The directory that is the mount point for the filesystem. This field is
identical to the Mount Point field when changing the logical volume
attributes.
To change the mount point of the filesystem, enter a new directory. The
directory specified is created (if it does not already exist) when the
Chapter 18. Changing the Volume Group, Logical Volume and Filesystem Attributes
18-13
filesystem is created. If you are installing the system, the data for the
filesystem is restored to the new mount point. For example, if the old
mount point was /home/files and the new mount point is to be /data/files,
then a file previously called /home/files/userinfo is restored as
/data/files/userinfo.
If changed, the /etc/filesystems file is updated to reflect the new mount
point for the filesystem so that all future attempts to mount the filesystem
will do so to the new mount point.
Fragment Size
The current filesystem fragment size, or fragsize. You can indicate a new
fragsize in bytes, and valid values are 512, 1024, 2048, and 4096. If you are
using filesystem compression, the fragsize must be set to 2048 or less.
A filesystem containing many small files can benefit from using a small
fragsize, as each file requires less space. Larger files with small fragsizes
will require more I/O requests and more system processing to read and
write larger pieces of data.
Note: If you increase the current fragment size value, a filesystem with
many small files might require more filesystem space, because each
individual file requires more physical disk space. Increasing this
value can cause the filesystem to fill when restoring the files. If the
filesystem is nearly full, increase the size of the logical volume for
this filesystem to provide added space.
Number of Bytes Per Inode (NBPI)
The current NBPI. The NBPI indicates the number of bytes of filesystem
space that accounts for each inode in the filesystem inode table.
Enter a new value, which must be 1024, 2048, 4096, 8192, 16384, or 32768.
The valid values differ depending on the value in the Allocation Group
(AG) size field. The valid values are shown at the bottom of the screen.
Note: Decreasing the NBPI value might cause a nearly full filesystem to run
out of space when restoring a large amount of data, because the
filesystem control information uses more space on the disk.
Increasing the NBPI value uses less space for filesystem control
information, but it is possible to run out of filesystem Inodes
(assigned to each file) when restoring a large number of files.
Use Data Compression?
Indicates whether or not the data is to be compressed in the filesystem.
Enter “y” or “n” to change the current value. Compressing the data
requires less disk space, usually between 25% and 50% depending on the
type of data. However, each read or write of data requires much more
system processing to compress and decompress the data.
Note: If you select not to use compression on a previously compressed
filesystem, the data uses more space in the filesystem as it is being
restored. Because it is not possible to estimate the amount of added
space required, change the logical volume size for this filesystem to
at least twice the Minimum size, indicated at the top of the screen.
Large Files Support?
Indicates whether large files are supported for this filesystem. Only if this
field is set to “yes” can you create files larger than 2 gigabytes in size
within this filesystem. This field appears only on AIX 4.2 or later systems.
18-14
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Enabling this support does not affect I/O performance, but the filesystem
cannot be remote mounted to another system that does not support this
feature.
To change the current value, enter “y” or “n”.
Allocation Group (AG) Size
Indicates the current allocation group size. The allocation group is a
grouping of inodes and disk blocks similar to BSD cylinder groups. This
field appears only on AIX 4.2 or later systems.
You can enter a new value in megabytes, which must be 8, 16, 32 or 64.
The valid values will differ depending on the value in the Number of
Bytes Per Inode (NBPI) field. The valid values are shown at the bottom of
the screen.
Chapter 18. Changing the Volume Group, Logical Volume and Filesystem Attributes
18-15
18-16
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Appendix A. Commands
© Copyright IBM Corp. 2002
A-1
cfgremrootaccess
Purpose
Adds, removes, or lists commands enables for Remote Command Access.
Syntax
cfgremrootaccess -h “host”
-d “host”
-l
-c “command”
-r
Description
The cfgremrootaccess command is used to configure a backup client to allow a
server to remotely initiate its backups. This is called a “pull backup.” For example,
A server contacts a client to start the backup; in response the client initiates the
backup and sends the data to the server. Therefore, the server is actually “pulling”
the backup operation from the client.
Using pull backups enables you to centrally manage your backups by initiating
and controlling backup operations from a single location. You can further automate
this central management of backups using the SysBack Scheduling and Scripting
functions described in Chapter 7, “Scheduled Backups and Scripts”, on page 7-1.
Pull backups require not only Remote Services Configuration, but also Remote
Command Access configuration. The Remote Services Configuration enables data
to be sent across the network to the backup server. The Remote Command Access
enables the server to initiate the pull backup.
Also note that the server that initiates the pull backup can be a different server
than where the data is sent. Because of this, Remote Services can be configured
between machine A, the client, and machine B, the backup server receiving the
backup, while Remote Command Access is configured between machine A, the
client, and machine C, the backup initiation server.
Flags
–c “command”
Specifies the fully-qualified path and name of a command to allow for remote
access. Used with the -h and -r flags.
–d“host”
Specifies the hostname of a machine that is not allowed to contact the client to
initiate a backup. Used only when the -h flag is set to ″ALL.″
–h “host”
Specifies the hostname of a machine that will initiate a backup of this client.
This flag is required and must be used with the -c flag.
–l Lists all commands enabled for Remote Command Access.
–r Indicates to remove the command specified with the -c flag.
A-2
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Examples
1. To allow the server sysback1.aix.dfw.ibm.com to contact this client to initiate a
pull backup using the /usr/sbin/sysback command, type:
cfgremrootaccess -c ’/usr/sbin/sysback’ -h ’sysback1.aix.dfw.ibm.com’
2. To list all commands that are enabled for Remote Command Access and their
respective hosts, type:
cfgremrootaccess -l
3. To remove the access for the /usr/sbin/mkdirback command, to be remotely
initiated, type:
cfgremrootaccess -r -c ’/usr/sbin/mkdirback’
Files
/usr/lpp/sysback/.root_access
Lists commands and respective hosts allowed
Remote Command Access.
Related Information
The cfgremsvs command.
Appendix A. Commands
A-3
chimagefile
Purpose
Changes the permissions of an existing disk image backup file.
Syntax
chimagefile
-l
-f ″filename″
-d ″description″
-h ″host″
-r
host =
h
a
user =
u
a
-v
-H ″hostname″
-u ″username″
Description
When you perform a backup to a disk image file, you can set the read permission
of the file to allow only the original host or original user to read from the file. You
can change this permission at a later time if, for instance, you need to use an
image file with “original host only” permission needs as a network installation
image for other systems.
Using the chimagefile command, you can change the owner of the file (host or
user), the backup description, or the host or user read permission of an existing
image file.
Only the following users can change the read permission of a disk image file:
v Root user on the system where the file physically resides
v Root user on the system from which the file originated
v User who originally wrote the file
A file can only be accessed by the root user or by the same user who created the
file if the file is in a directory the user has permission to access as defined by the
cfglocalaccess or cfgremaccess commands. Only the root user can read backup
image files contained in directories not explicitly defined.
Flags
–d “description”
Changes the user-defined backup description to the specified string. This
description can be up to 60 characters. A longer string will be truncated to 60
characters. The description must be within double quotation marks (″) and
cannot contain single or double quotation marks.
–f “filename”
Changes the filename.
–h “host”
Changes the filename on the specified host.
–l Lists the current values, separated by colons (:), in the following order:
A-4
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
host_owner:user_owner:host_perm:user_perm:description
This flag is normally used only by SMIT for querying the current values.
–r host=“[h/a]”
Changes the host read permission to the specified value. The possible values
are h, indicating that only the host that owns the file can read it, or a,
indicating that any host can read the file. The root user on the server where the
file physically resides can also access the file.
–r user=“[u/a]”
Changes the user read permission to the specified value. The possible values
are u, indicating that only the user that owns the file can read it, or a,
indicating that any user can read the file. The root user on the server or the
root user on the host that owns the file can also access the file. Note that if any
host is allowed to access the file, but only the user that owns the file can read
it, the user who owns the file can access it from any host.
–v
Report the old and new values when they change. Without this flag, there is
no output from this command.
–H “hostname”
Changes the current owner of the file to host hostname. Note that, if the host
read permissions are currently set to this host only, changing the owner of the
file to a new host prevents the original host from accessing the file. The root
user on the server where the file physically resides can always access the file.
–U “username”
Changes the current user owner of the file to username. Note that, if the current
user permission is set to this user only, changing the owner of the file to a new
user will prevent the original user from accessing the file. The root user on the
host that owns the file, or the root user on the server in which the file
physically resides will always be able to access the file.
Examples
1. To change the image file /usr/lpp/sysback/bf/pluto/root/plutoV.220934 to allow
all users on all hosts to read this file, type:
chimagefile -r host=a -r user=a -f \
/usr/lpp/sysback/bf/pluto/root/plutoV.10220934
2. To change the current host of the file
/usr/lpp/sysback/images/anyhost.uvervg.043141022 to host mercury, and the
backup description to master uservg volume group, assuming the file exists on
the server mars, type:
chimagefile -d "master uservg volume group" -H mercury \
-h mars -f \
/usr/lpp/sysback/images/anyhost.uservg.043141022
Related Information
The cfglocalaccess and cfgremaccess commands.
Appendix A. Commands
A-5
chinstclient
Purpose
Sets client defaults for use during a tape or network boot/installation process.
Syntax
chinstclient -b
N
T
-h ″host″
-M
prompt
noprompt
-D
-M ″console device name″
-t ″term type″
-v
root
all
import
-m ″key″
-r
-l
-d ″filename″
-i ″number″
-R
y
n
-N
y
n
-S ″ip address″
-g ″ip address″
-n ″device name″
-s ″subnet mask″
-T ″value″
-E
en
et
-e
bnc
dix
-p ″post install script file″
-L ″LVM information files″
-o
Description
The chinstclient command enables you to specify response to common prompts
during a boot and installation process. Setting these defaults can minimize, and in
some cases eliminate, the need for user prompting during a boot and installation
process.
These defaults are set on a client-by-client basis.
Flags
-b “N | T”
Specifies if the defaults are for a tape or network boot and installation process.
-h “host”
Specifies the name of the client.
-M “prompt | noprompt”
Specifies whether or not to prompt the client. If this field is set to “noprompt”,
A-6
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
the installation menus appear after a network boot. Otherwise, the installation
occurs after the network boot without prompting the user, as long as all
required information for the user is also provided.
-D
Turns on debug logging operations for troubleshooting network boot and
installation problems.
-c “console device name”
To prevent the prompt before installation that determines the console to be
used for installation, enter the name of the console here. The console can be
either an LFT (graphical display) or TTY (ascii display) device. Examples are
“/dev/lft0” or “/dev/tty0”. If an LFT device exists on the client, you can
select it by typing /dev/console.
Note: You must select a display, even for no-prompt installations, in order to
receive error or warning messages and to enable prompts if the default
information provided is insufficient.
-t “term type”
If you specified a TTY device, specify a terminal type. This ensure that the
installation menus, if required, display with the correct terminal emulation.
Examples of common terminal types are “ibm3151”, “tvi925”, and “vt100”.
-v “root | all | import”
Specifies how to handle volume group processing during installation. The
available options are:
Recreate and restore rootvg only
Only the rootvg volume group is created and restored from the media,
even if other volume groups are defined on the backup. At the end of the
installation, rootvg is the only defined volume group.
Recreate and restore all volume groups
All volume groups included on the backup media are created and restored.
Volume groups defined on the original system but whose data was not
included on the backup, are created but no data is restored.
Recreate and restore rootvg and import all other volume groups from disk
The rootvg volume group is created and restored and all other volume
groups are imported from disks on the client. This options assumes that
the non-rootvg volume groups already exist on the client. This option is
typically used when restoring the rootvg volume group after a system
failure where no other volume groups were affected.
-d “filename”
Specifies the name of the installation image file that contains the backup image
to be used or an installation tape device. If you are installing from a file, the
directory containing the file was configured using the cfgremaccess command
and is the only directory path available for installation image files. Specify the
file here to prevent being prompted for it during installation.
-i “number”
If the client is to be installed from tape and the system backup is not the first
backup on the tape, enter the backup sequence number. This number must be
between 1 and the total number of system backups on the tape.
-R “y | n”
Indicates whether or not to remove network specific information required to
Appendix A. Commands
A-7
configure the host to the network. This option is useful when the machine
installed was cloned from a backup image of a machine that is still present on
the network.
-S “ip address”
The address of the installation server. If the backup media from which the
client will be installed exists on a different server than the network boot server,
enter the IP address of the installation server here.
-g “ip address”
The IP address of the gateway, if the client will be installed from a different
server than the network boot server, and a different gateway address is
required to reach the installation server.
-n “device name”
The network device name, if the client will be installed using a different
network adapter than it was booted from. The name should be in the format of
“entX”, or “tok0”.
-s “subnet mask”
The subnet mask required for the network adapter used to contact the
installation server, if the client will be installed from a different network
adapter than it was booted from.
-T “value”
The token ring speed of the network for network installations over a token
ring interface. If the boot server is also the installation server, you do not need
to set this option. If the installation server uses a different network adapter,
and the network adapter is token ring, specify the correct token ring speed.
-E “en | et”
The ethernet interface used for network installations over an ethernet interface.
If the boot server is also the installation server, you do not need to set this
option. If the installation server uses a different network adapter, and the
network adapter is ethernet, specify the correct interface.
-e “bnc | dix”
Specifies the ethernet connection type for network installation over an ethernet
interface. If the boot server is also the installation server, you do not need to
set this option. If the installation server uses a different network adapter, and
the network adapter is ethernet, specify the correct connection.
-p “post install script file”
The fully-qualified path and name of a post-install script to run at the end of
installation. The file must exist on the installation server and be readable by
any user. For more information about post-install scripts, see Appendix C,
“Creating Scripts for Customizing the System Backup and Install Process”, on
page C-1.
-L “LVM information file”
The fully-qualified path and name of an LVM information file. The LVM
information file must contain information that is compatible with the client.
The file must exist on the installation server and be readable by any user. For
more information about LVM information files, see Chapter 15, “Utilities”, on
page 15-1.
-o Overrides the /etc/SDR_dest_info file from the boot server. If a system backup
image was created on an SP node from a different SP complex, the
/etc/SDR_dest_info file will refer to the wrong control workstation, and
A-8
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
post-installation customization will fail. Specifying this options causes the
correct file to be copied from the boot server before the pssp_script post-install
script runs.
Appendix A. Commands
A-9
cfglocaccess
Purpose
Adds, lists, or removes local user access to backup devices and directories.
Syntax
cfglocaccess
-l
-u ″user″
-r
-d
-f
-b
-l
″user″
″device″
″directory″
″directory″
Description
All users on the system are provided access to all of the backup devices on the
system when SysBack is first installed. This is accomplished by providing a
permission record for a user called all. You might not want certain users to have
access to all devices and backup directories, so you can deny specific users access
or change the device or directory list that specific users are allowed to use. The
devices and directories specified when using this option determine the list of
backup options that users receive when they perform backups using SMIT, and
those devices or directories that other SysBack commands, executed at the
command line, are permitted to read or write to.
If you do not want all users on the system to have the same access, remove access
to user all, using the -r flag, before setting up access to specific users.
SysBack provides flexibility in assigning users access to devices, directories, and so
on. For instance, you can assign specific devices or directories to each user on the
system or assign a device to all users on the system except specific users. Refer to
“Adding or Changing Local User Device/Directory Access” on page 8-2 for details
and examples of user permissions.
Flags
-b “directory”
Specifies one or more directories, separated by spaces, to which the user can
perform regular backups. The user can back up to any subdirectory in the
directories provided here. When performing any SysBack backup, the user has
a disk image file backup option for each directory specified.
-d “user”
Specifies users, separated by spaces, that are to be explicitly denied access. If
you use this flag when you are adding or changing a permission record for all
users, all users except those specified will be permitted.
-l
Lists current user permissions.
-f “device”
Specifies a list of tape drives or virtual devices, separated by spaces, to which
the user or users can perform backups. Only the devices entered here are
displayed on the SMIT menus when the user selects a backup.
A-10
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
-i “directory”
Specifies one or more directories, separated by spaces, to which the user can
perform network installation image backups. This field applies only to the root
user, because only the root user can create network installation image backups
of volume groups. This flag applies only when you configure the root user or
all users, if root is not specifically defined.
When backing up a volume group to use as a network installation image, the
user has a disk image file backup option for each directory specified. Also,
when this or other systems attempt to perform a network installation, all
backup images in the specified directories that the client system has permission
to read will be displayed as installation options.
-r
Removes the specified user access record.
-u “user”
Specifies the user to be added, changed, or deleted.
Examples
1. To add user peter to the access list, providing access to local tape device
/dev/rmt0, virtual device vdev0, and the backup directory
/usr/lpp/sysback/bf/local/peter, type:
cfglocaccess -u peter -d "/dev/rmt0 vdev0" -b \
/usr/lpp/sysback/bf/local/peter
2. To allow all users, except users marsha and jan, access to all devices, type:
cfglocaccess -u all -d "marsha jan" -f all
3. To remove the access record for user cindy from the access list, type:
cfglocaccess -r -u cindy
4. To list current hosts allowed to access local backup device access, list:
cfglocaccess -l
Files
/usr/lpp/sysback/.local_access List of local users allowed or denied access to the
backup devices and directories for all SysBack
backups. The format of the data in this file is very
specific and should not be edited manually.
Related Information
The cfgremserver command.
“Adding or Changing Local User Device/Directory Access” on page 8-2.
Appendix A. Commands
A-11
cfgremaccess
Purpose
Adds, lists, or removes remote host and user access to local backup devices and
directories.
Syntax
cfgremaccess
-l
-u ″user″
-r
-d
-D
-f
-b
-l
″host″
″user″
″device″
″directory″
″directory″
Description
Before any client host can perform a backup to a remote server, the server must
enable permission to the client host and to the specific users on the client who will
be performing the backup operations. The cfgremaccess command provides access
to the specified hosts and users. To provide the remote host a list of valid devices
and backup files on this server when executing backups from SMIT, you must also
run the cfgremserver command on the client to identify this server.
Before using this command, you must have already configured this host for
Remote Services using the cfgremsvs command. You must also run the cfgremsvs
command on the client host before attempting to execute any backups to this
server.
You can provide all hosts on the network access to the server by creating a single
record that applies to all of the hosts by providing a permission record for a host
called all. You can exclude or deny specific hosts by using the -d flag. You can also
create an access record for specific hosts, to override the all record for that host
only.
Likewise, if you want to deny access to most, but not all, users on a particular
host, you can indicate a user name of all. You can then supply a list of specific
users to deny with the -D flag. You can also create a permission record for a
specific host and user, to override any records for all hosts or all users.
Refer to “Adding or Changing the List of Remote Backup Servers” on page 9-7 for
a complete explanation of the permissions and example of their use.
Note: You can enter either the hostname or a full domain name, if any. If you are
operating in a domain network, use the full domain name of the client host.
Using only the host name in a domain network allows access to any host on
the network with the given host name, even though they might be in
different domains.
Flags
-b “directory”
Specifies one or more directories to which the client hostname and username
can perform regular image file backups. The client can perform a backup to
A-12
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
this directory or any of its subdirectories. A backup image file option for each
directory supplied is displayed in SMIT when the user attempts to perform
any backup, list, verify or restore operation.
-d “host”
Denies access to one or more specified hostnames. This flag is valid only when
all is entered for the hostname (-h) parameter.
-f “device”
Specifies one or more devices to which the client hostname and username can
perform regular image file backups. The entry can be a tape device (such as
/dev/rmt0), virtual device (such as vdev0) or all, which allows the client to
back up to any devices on this server. A device option for each device supplied
is displayed in SMIT when the user attempts to perform any backup, list,
verify, or restore operation. Also, these devices are listed when the client
attempts to perform a network installation from this server.
–h “hostname”
Specifies the hostname of the client to add or remove. A hostname of all allows
access to all hosts on the network. With a hostname of all, you can also enter
hosts to exclude with the -d flag.
–i “directory”
Specifies one or more directories to which the client hostname and username
can perform network installation image file backups. The client can perform a
backup to this directory or any of its sub-directories. Any files in the
directories supplied will be listed for the client when it attempts to perform a
network installation operation.
-l
Lists all hosts currently configured, along with the devices and directories to
which they have access.
-u “username”
Specifies the username on the client to add or remove. To allow access to all
users on the specified hostname, enter the username all.
-r
Removes the access record for the specified hostname and username.
Removing the access record does not necessarily deny access to the client,
because it might revert to access provided by another record.
-D “user”
Denies access to one or more specified user names. This flag is valid only
when all is entered for the username (-u) parameter.
Examples
1. To add user peter on client host pluto to the access list, providing access to local
tape device /dev/rmt0, virtual device vdev0, and the backup directory
/usr/lpp/sysback/bf/pluto/peter, type:
cfgremaccess -h pluto -u peter -d "/dev/rmt0 vdev0" -b \
/usr/lpp/sysback/bf/pluto/peter
2. To allow all users on all hosts, except users marsha and jan on host mars, access
to all devices, type:
cfgremaccess -h all -d mars -u all -D "marsha jan" -f all
3. To remove the access record for user cindy on host venus from the access list,
type:
cfgremaccess -r -h venus -u cindy
4. To list current hosts allowed local backup device access, type:
cfgremaccess -l
Appendix A. Commands
A-13
Files
/usr/lpp/sysback/.remote_access
List of hosts and users allowed or denied access to
the local backup devices and directories for all
SysBack backups. The format of the data in this file
is very specific and should not be edited manually.
Related Information
The cfgremsvs and cfgremserver commands.
“Adding or Changing the List of Remote Backup Servers” on page 9-7.
A-14
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
cfgremserver
Purpose
Defines, lists, or undefines servers the client can access.
Syntax
cfgremserver
-l
-a
-r
-h hostname
-f
Description
The cfgremserver command is used to either define or undefine a remote host that
is to act as a server for the local host. The local system can execute remote sysback
commands without setting up the server with this command, but only devices on
remote servers set up with this command will be listed when executing backup,
list, verify or restore options from SMIT. This command can also be used to list
currently defined remote servers.
Before this command can be used to define a server to the system, both the local
host and the host to act as the backup server must be configured for Remote
Services using the cfgremsvs command. In addition, before a client machine can
add a remote server, that server must have already provided access to its devices
or directories with the cfgremaccess command. When defining a new remote
server, this command verifies that Remote Services is configured on both the client
and server systems and the server has granted access to this client.
Flags
-a Adds the hostname to the list of remote servers.
-f
Indicates whether or not remote disk image file backup options should be
displayed from SMIT. Without this option, disk image file options are not
listed as valid backup devices even if the server has enabled directory access to
this client.
-h “hostname”
Hostname of host where the requested backup device is attached.
-l
Lists defined servers. Searches each server in the list to obtain a list of all
devices and directories accessible by this client.
-r
Removes the hostname from the list of remote servers.
Examples
1. To add server venus to the list and provide backup options on venus for
devices or disk image files, type:
cfgremserver -a -h venus -f
2. To remove the server venus from the list of remote servers, type:
cfgremserver -r -h venus
3. To list currently defined servers and the devices available on those servers,
type:
cfgremserver -l
Appendix A. Commands
A-15
Files
/usr/lpp/sysback/.servers
List of servers for which to list SMIT backup
options
Related Information
The cfgremsvs and chremaccess commands.
A-16
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
cfgremsvs
Purpose
Configures Remote Services.
Syntax
cfgremsvs
Description
This command is used to configure Remote Backup Services for either a host that
is to act as a backup server for other machines or on a client that requires access to
backup devices located on another backup server.
The command requires no arguments, and when executed, creates a new user ID
“sbnet” in the staff group. This user ID is used to execute commands and transfer
backup data using the sbclient command from the client to the server machine.
You need to run this command only once on each workstation.
Files
/etc/password
User password file
/etc/group
User group file
/usr/lpp/sysback/.remote_access
Access permissions for remote hosts
Related Information
The sbclient command.
Appendix A. Commands
A-17
cfgvirdev
Purpose
Adds, lists, or removes virtual device descriptions.
Syntax
cfgvirdev
-l
-r
-a
device
-s
-p
-c
-A
-d ″description″
Description
The cfgvirdev command is used to add, list, or remove a virtual device description
from the system. A virtual device can have a write policy of either sequential,
parallel or multi-copy. Refer to Chapter 10, “Virtual Devices”, on page 10-1 for a
detailed description of each write policy.
You can also indicate that a sequential autoloader is attached to the virtual device.
This indicates that when a tape reaches the end of media, it is ejected from the
drive, and the reading or writing continues automatically as soon as the next
volume is inserted. In this case, the user receives no prompt to change the volume.
If you are creating an autoloader virtual device, you can assign one or more
physical devices. Otherwise, you must assign multiple physical devices. A device is
a tape drive such as “/dev/rmt0”.
You can assign your own description to the virtual device, or you can allow the
system to provide a default description. By default, the description assigned will
be either “Sequential Virtual Device”, “Parallel Virtual Device” or “Multi-copy
Virtual Device”, depending on the write policy you choose.
When adding a new virtual device, the device option can be a single tape drive
(such as “/dev/rmt0”) or a list of tape drives (such as “/dev/rmt0
/dev/rmt1/dev/rmt2”). When removing a virtual device, the device option is the
virtual device name to remove. Omit the device option when only listing defined
virtual devices.
When adding a new virtual device, the system automatically assigns the next
available virtual device name. The prefix will be vdev, followed by a device
number, starting with 0 and increasing for each subsequent device added.
Flags
-a Adds a new virtual device.
-r
Removes an existing virtual device.
-l
Lists defined virtual devices.
-s
When adding a virtual device, indicates a Sequential write-policy.
-p
When adding a virtual device, indicates a Parallel write-policy.
A-18
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
-c
When adding a virtual device, indicates a Multi-copy write-policy.
-d “description”
The description of the virtual device. This description is displayed on the SMIT
screens when listing available backup and restore devices. Enclose this
description in double quotation marks.
-A
Indicates an autoloader is attached to the physical devices.
Examples
1. To create a sequential virtual device, using tape devices rmt0 and rmt1, type:
cfgvirdev -as /dev/rmt0 /dev/rmt1
2. To create a parallel virtual device using devices rmt1 and rmt2, and assigning a
user description, type:
cfgvirdev -apd "Parallel 8mm tapes" /dev/rmt1 /dev/rmt2
3. To define rmt0 as an auto-loader, type:
cfgvirdev -asd "Autoloader" /dev/rmt0
4. To remove virtual device vdev0, type:
cfgvirdev -r vdev0
Appendix A. Commands
A-19
chexcludelist
Purpose
Adds, removes, or lists files or directories in the exclude list file.
Syntax
chexcludelist
-l
-r
-a
″filename″
″@LVname″
″filename″
″@LVname″
Description
The chexcludelist command is used to either add or delete entries from the
exclude list file. The exclude list file contains directories, individual filenames, or
logical volume names that are to be excluded from all backups created with the
SysBack commands.
The file or directory name must begin with a slash (/) but can contain wildcard (*)
characters. If you enter a filename, only that file is excluded. If you enter a
directory name, all files and directories within the specified directory are excluded.
When specifying wild cards, you must enclose the string in single quotation marks
(’) to preserve the special characters.
The following are examples of using wildcard in the exclude list:
/tmp
All files below the /tmp directory
/*test*
All files in the system containing the word “test”.
/*old
All files on the system ending with “old”.
/home/b*
All files under /home starting with a “b”.
/home/t*y
All files under /home starting with “t” and ending with “y”.
Note that any entry starting with “/*” will include all directories on the system,
not just the root (/) directory.
Note: When excluding a logical volume name, type a “@” character before the
logical volume name. This indicates that the entry is a logical volume and
not a file or directory. You cannot use wild cards in logical volume names.
Flags
-l
Lists current entries. No changes are made to the existing exclude list.
-a “filename”
Adds a filename or directory to the exclude list.
-r “filename”
Removes a filename or directory from the exclude list.
A-20
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Examples
1. To add the directory /tmp to the exclude list, type:
chexcludelist -a /tmp
2. To exclude all files on the system ending with “.old”, type:
chexcludelist -a ©/*.old©
3. To exclude logical volume userlv1 from volume group backups, type:
chexcludelist -a @userlv1
4. To remove the file /home/test from the exclude list, type:
chexcludelist -r /home/test
5. To list current excluded files and directories, type:
chexcludelist -l
Files
/usr/lpp/sysback/.exclude_list Filenames and directories to exclude from backups.
Appendix A. Commands
A-21
editlvminfo
Purpose
Customizes an LVM information file.
Syntax
editlvminfo -f ″filename″
Description
The editlvminfo command reads the contents of the LVM information file, specified
by the filename parameter and initializes the user interface used for changing the
volume group, logical volume, and filesystem information within the file.
This option is usually used to customize a file, created with the mkvginfo
command, before including the customized file on a backup, using the -g flag with
either the mklvback, mkjfsback, mkvgback or sysback commands.
The user interface provides options for changing any of the information in the file,
such as assigning new physical volumes to volume groups, changing the volume
group where a logical volume or filesystem exists, and changing the sizes of logical
volumes and filesystems.
Detailed instructions for changing the information in the LVM information file is in
Chapter 18, “Changing the Volume Group, Logical Volume and Filesystem
Attributes”, on page 18-1.
Related Information
The mkvginfo, mklvback, mkjfsback, mkvgback or sysback commands.
Chapter 18, “Changing the Volume Group, Logical Volume and Filesystem
Attributes”, on page 18-1.
A-22
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
getlvminfo
Purpose
Displays Logical Volume Manager (LVM) information for volume groups, logical
volumes, and physical volumes.
Syntax
getlvminfo
-V
-l
-P
-G
Option
Option
Option
Option
-V Option:
-V ″VGname″
-c
-d
-i
-l
-q
-x
-A
-S
-1
-a
-b
-c
-d
-e
-i
-l
-w
-x
-z
-X
-Z
-1
-a
-l
-n
-v
-x
-A
-S
-1
-p I
-p N
-p X
-l Option:
-l ″LVname″
-p
-m
-I
-N
-X
-h
-P Option:
-P ″PVname″
-I
-N
-X
-G Option:
-G ″PVname″
-c
-d
-i
-l
-s
-X
-1
-p I
-p N
-p X
-I
-N
Appendix A. Commands
A-23
Description
This command is used to display all available Logical Volume Manager information
for volume groups, physical volumes, and logical volumes. It retrieves LVM
information for all SysBack commands requiring LVM data. The getlvminfo
command is faster and easier than using the standard AIX LVM commands and
provides the information in a simple, reliable format.
At least one argument is required. If only a single argument (-V or -P) is specified,
all volume groups or physical volumes on the system, respectively, are displayed.
Any time the -I, -N or -X arguments are used, any physical volumes are displayed
using the PVID, PV Name or PV Location, respectively.
One of the flags -V, -P, -L or -G must be specified to display attributes for a
volume group, physical volume, or logical volume. The -G flag can also display
volume group information for a volume group that is not currently varied on.
Flags
–1
Displays all output on a single line, separated by spaces. The default is to
show each attribute on a separate line.
–a
When used with the –L flag, displays the intra-physical volume policy for a
logical volume. When used with the –P flag, displays currently allocated
physical partitions for a physical volume.
–b
Used with the -L flag to display whether or not bad block relocation is enabled
for a logical volume.
–c
When used with -V or -G flag, indicates whether or not a volume group is
concurrent-capable. When used with -L flag, indicates the number of copies for
a logical volume.
–d
With the -V or -G flag, indicates the number of physical volumes assigned to
the volume group. When the -L flag, indicates a mirror scheduling policy of
either parallel (p) or sequential.
–e
Used with the -L flag, displays the intra-physical volume policy for a logical
volume.
–h
Displays a more detailed usage message for the getlvminfo command to
standard error.
–i With the -V or -G flag, displays the volume group ID (VGID); with the -L flag,
displays the logical volume ID (LVID); and with the -P flag, displays the physical
volume ID (PVID).
–l Used with the -V or -G flag to list all logical volumes names in the volume
group. With the -L flag, indicates the logical volume label.
–m
Used with -L flag to list a physical partition map for the logical volume. This flag
cannot be used with the -p flag.
A-24
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
–n
With the -L flag, indicates the number of logical partitions (LPs) assigned to the
logical volume. With the -P flag, displays the total physical partitions (PPs)
available on the physical volume.
–p
With the -V or -G flag, displays the physical volumes in the volume group.
With the -L flag, displays the physical volumes on which a logical volume
resides. This flag cannot be used with the -m flag.
–q
Used with the -V flag to indicate whether or not quorum checking is enabled for
a volume group.
–r Used with the -L flag to indicate whether or not the logical volume is
relocatable.
–s
Used with the -L flag to indicate whether or not a strict allocation policy is
enabled for a logical volume, preventing a single copy from being placed on
multiple physical volumes.
–t Used with the -L flag to display the logical volume type.
–u
Used with the -L flag to display the upperbound limit (maximum PVs) for the
logical volume.
–v
With -L flag, displays whether or not write-verification is enabled. With the -P
flag, displays the number of volume group descriptor areas (VGDAs) on the
physical volume.
–w
With -L flag, displays whether or not mirror-write-consistency checking (MWC) is
enabled.
–x
With the -V flag, indicates whether or not the volume group is varied on in
concurrent mode. With the -L flag, indicates the maximum logical partitions (LPs)
for the logical volume.
–z
With the -L flag, indicates the logical volume stripe size. A value of zero (0)
indicates that striping is disabled.
–A
With the -V flag, indicates whether or not the volume group is automatically
varied on at system startup.
-G “PVname”
Displays volume group information using the physical volume name instead of
the volume group name.
–I Whenever physical volume information is displayed, this flag indicates that the
physical volumes should be displayed using the physical volume ID (PVID).
This is the default.
–L “LVname”
Displays attributes for the logical volume name specified.
–N
Whenever physical volume information is displayed, this flag displays physical
Appendix A. Commands
A-25
volumes using the physical volume name (hdiskX) rather then the PVID. When
displaying a list of volume groups using the -V flag, this flag displays the
volume groups by name rather than VGID.
–P “PVname”
When used without specifying a physical volume name, this flag displays a list
of physical volumes on the system. When a physical volume name is specified,
this flag displays the attributes for the specified physical volume.
–S With -V or -G flag, displays the physical partition size in megabytes. With the -L
flag, displays the logical volume state as either syncd (0) or stale (1). With the -P
flag, displays the physical volume state of either active (0), removed/missing (1),
or stale (2).
–V “VGname”
When used without specifying a volume group name, this flag displays a list
of volume groups on the system. Use with the -N flag to display the volume
group list by volume group name instead of VGID. When a volume group
name is provided, this flag displays the attributes for the specified volume
group.
–X Any time physical volumes are displayed with this command, displays them
by location code rather then by PVID.
–Z
Used with the -L flag to indicate the stripe width (number of physical volumes
used in striping) for the logical volume. If a value of zero (0) is returned,
striping is disabled.
Examples
1. To list the logical volumes in the rootvg volume group:
getlvminfo -V rootvg -l
2. To list the Intra-PV policy, Inter-PV policy, number of copies and number of
logical partitions for the userlv1 logical volume on a single line:
getlvminfo -L userlv1 -aecn1
3. To list the physical volumes in the uservg volume group by location code:
getlvminfo -V uservg -pX
4. To show the physical partition map for the datalv2 logical volume in the form
“hdiskname:PPnumber”:
getlvminfo -L datalv2 -mN
A-26
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
lscfginfo
Purpose
Prints or displays a report of the Logical Volume Manager configuration.
Syntax
lscfginfo
-f ″filename″
-q ″queue″
Description
This command prints, displays, or saves to file a report containing the system
configuration. This includes all information pertaining to the volume groups,
physical volumes, logical volumes, and filesystems. This report contains all of the
key information that might be necessary for rebuilding various components of the
system in case of system failures.
In most cases, it is not necessary to use this information to recover from system
failures, because the SysBack installation process and the remakevg command
recreate volume groups, logical volumes, and filesystems without requiring the
user to have knowledge of the system configuration. This information, however,
can be useful in planning a migration of filesystems to new hardware or in
reorganizing an existing system to take advantage of disk striping, mirroring, and
other features.
The report output includes the following information:
Volume group information:
One record for each volume group. Contains all volume group attributes and
physical volume sizes. Example:
Volume Group
-----------rootvg
Auto-on?
-------y
PP Size Quorum? Total
Used
Free
PV(s)
------- ------- -------- -------- -------- -------4 MB
y
250 PPs
213 PPs
37 PPs hdisk0
1002 MB
852 MB
148 MB
Logical Volume Information:
One record for each logical volume. Contains most logical volume attributes
and a list of the physical volumes where each logical volume resides. Example:
Volume Group
Logical Volume
Attributes
-------------- ---------------- -----------------------------------------rootvg
hd8
Type:
jfslog
Copies: 1
Upper: 32
LPs:
1
IntraPV: c
StrpSz: 0
MinLps: 0
InterPV: m
PV(s):
hdisk0
Physical Volume Information:
One record for each physical volume. Contains the physical volume location,
size, and a map of the regions on the disk where each logical volume resides.
Example:
PV Name Volume Group
Location
PVID
MB
PPs
-------- ------------- ------------ ----------------- ----- -----hdisk0
rootvg
00-00-0S-0,0 00000218acd0f607 1002 250
LV Name
PPs
Region
-------------------------------hd8
101
|....X.....|
hd6
51-62
|..X.......|
Appendix A. Commands
A-27
hd4
hd3
hd2
73-96
3
102-103
105-109
5-27
110-207
222-242
|..XX......|
|X.........|
|....X.....|
|....X.....|
|XX........|
|....XXXXX.|
|.........X|
Filesystem information:
Contains all filesystem attributes. Note that the BF Spt and AG Size fields are
displayed only on AIX 4.2 or later systems.
Example:
Volume Group
Mount Point
-------------- ---------------rootvg
/
/home
/tmp
/usr
/var
/data/files
FragSz
-----4096
512
4096
4096
512
512
NBPI Cmprs BF Spt? AG Size?
---- ----- ------- -------2048
no
n
8
4096
no
y
8
4096
no
n
8
4096
no
n
8
4096
no
n
8
4096
no
y
32
Logical Volume Partition Maps:
Shows the specific physical partitions on the physical volumes that are used by
each logical volume. This is helpful in determining the fragmentation of logical
volumes.
Example:
Volume Group
Logical Volume Copy# PV Name
-------------- -------------- ----- --------rootvg
hd8
1
hdisk0
hd6
1
hdisk0
1
hdisk0
hd4
1
hdisk0
1
hdisk0
hd3
1
hdisk0
hd2
1
hdisk0
1
hdisk0
1
hdisk0
PP#s
------101-101
51-62
73-96
3
102-103
105-109
5-27
110-207
222-242
PV Region
-----------|....X.....|
|..X.......|
|..XX......|
|X.........|
|....X.....|
|....X.....|
|XX........|
|....XXXXX.|
|.........X|
Flags
-f “filename”
Saves the report in the named file. When used, no information is displayed on
the screen.
-p “queue”
Sends the report to the named print queue. When used, no information is
displayed on the screen.
Related Information
The getlvminfo command.
A-28
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
mkdirback
Purpose
Backs up specified files or directories.
Syntax
mkdirback
-v
-x
-b
-c
-d
-e
-h
-i
-m
-n
-O
-p
-r
-U
″size″
″termdev″
″description″
-f ″device″
-f ″directory″
-f ″file″
-I ″ID″
″hostname″
″days″
″host″
directory or file
Description
The mkdirback command is used to back up specific files or directories. You must
choose a starting directory to back up that directory and any sub-directories.
You can specify a single directory or multiple directories to include all of those
directories in a single backup. You can also specify specific filenames to back up
only those files. Filenames must be specified relative to the root directory by
preceding the filename with a slash (/).
The backup can be performed to a tape drive, virtual device, or disk image file.
When specifying a disk image file, you can enter the full pathname of the file to
create or overwrite. When doing so, the filename must begin with the prefix “FD”.
to indicate the file is a file/directory backup. You can also specify only the backup
directory and a unique ID. When doing so, a filename is created for you. Refer to
“Backups to Disk Image Files” on page 2-6 for information on the filename created.
If you are creating a file and a file by the same name already exists, specify the -O
option to overwrite the existing file.
The backup can also be performed to a remote device, virtual device, or disk
image file. If Remote Services has been configured on both the local and server
system using the cfgremsvs command, and you have defined a backup device on
the server for this system using the cfgremaccess command, then you can also
select a hostname using the -h option. The backup will then be written to the device
or filename on the specified server, provided the server has enabled this system
access to the specified device or directory.
Files, or files within directories, that are specified in the exclude list are not backed
up with this command. See chexcludelist command for information on creating
exclude lists.
Appendix A. Commands
A-29
Unlike all other backup types, this type of backup contains no Logical Volume
Manager (LVM) information. Therefore, it is not possible to use this backup,
regardless of its contents, for recreating volume groups, logical volumes, and
filesystems. Using this backup option for backing up all files on the system (from
the / directory) cannot provide a backup that can be used to reinstall the system.
Some files and directories on the system, such as certain contents of the root (/)
and /usr filesystems, should never be restored from a backup, as this would
adversely affect the active system configuration and can cause either system errors
or a complete system failure. Use this command only to back up files and
directories that contain non-system-related information, such as user data,
application programs, and so on.
Using the -m option, you can also specify to back up only files that have been
modified within a certain number of days. This is useful for performing daily
backups. Be aware, however, that a day is considered exactly 24 hours. Therefore,
always back up at least every 24 hours when specifying the -m1 option. If, for
instance, your backups were performed 1/2 hour later today than yesterday,
neither today’s nor yesterday’s backup will contain any files changed within that
1/2 hour. It is a good idea to use -m2 each day to ensure that no files will be
skipped regardless of the time the backup is performed.
Always perform a backup of all files, regardless of date, before performing partial
backups using the -m option. If you later need to restore files or directories, restore
them from a full backup, and then restore each partial backup created after the full
backup.
The backup can also be performed as a “pull” backup to enable central
management of backups. The machine that “pulls” the backup is called the
“initiator,” while the machine that is backed up is called the “backup system.” The
machine that receives the backup data is called the “destination system.”
The initiator is the machine where the mkdirback command is issued. The backup
system is referenced in the command syntax by the -U “host” flag. The data
destination system is referenced by the -h “host flag”.
When only the -U “host” flag is specified, the -f “device” flag refers to a device on
the initiator system. In this case, the backup destination and the initiator are the
same system. This is known as a “two-way pull backup.”
When the -U “host” and -h “host” flags are specified, the -f “device” flag refers to
a device on the host specified by the -h “host” flag. In this case, the backup
destination and the initiator are different systems. This is known as a “three-way
pull backup.”
Remote Services must be configured, using the cfgremsvs command, on all of the
machines involved in the backup operation, the initiator, the backup system, and
the data destination machines. The backup destination machine must have the
Server Options of Remote Services configured to allow the backup system machine
to send its data to devices on the backup destination machine. You can do this
with the Add or Change Client Host Access to this Server option or the
cfgremaccess command.
A-30
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
The backup system machine must have the Client Options of Remote Services
configured to specify the backup destination machine as a backup host. You can do
this with the Add or Change List of Remote Backup Servers option or the
cfgremserver command.
The backup system must also configure Remote Command Access to allow the
initiator machine to pull the backup.
Flags
–b ″size″
Specifies the buffer size in K-bytes (1024-bytes). Changing this value either
increases or decreases the amount of data written to the output device at one
time. Some devices with small buffers might require you to reduce this value,
while other devices with very large buffers can benefit from increased
performance by increasing this value. If you are unsure, use the default value
(64 Kbytes).
-c “termdev”
If the backup spans multiple tape volumes, the volume prompt is sent to the
termdev device (such as /dev/tty0). If the -hhostname option is specified, this
flag specifies the device on the remote host.
-d “description”
A custom description to be included in the backup header. If the description
includes spaces, the entire description must be enclosed in double quotation
marks (″).
-e Forwards to the end of the last SysBack backup on the tape and is used when
stacking multiple backups on one tape.
-f “device, file or directory”
Specifies the output device name, filename, or directory for the backup. The
specified output option can be on the local system or server (if -h option
provided). Device name can be a tape drive or virtual device name. If you
specify a directory, you must also specify an ID using the -I flag.
-h “hostname”
Name of the host to receive the backup data (for example, the data destination
host). This is the host that provides the backup device specified using the -f
flag.
-I “ID”
This option is used only if a directory is specified as the output device. The ID
is included in the filename created to make it unique from other disk image
file backups. If a backup in the same directory with this ID already exists,
specify the -O flag to overwrite the previous backup.
-m “days”
Specifies that only files modified within a given number of days will be backed
up.
-n
Indicates that the tape should not be rewound at the beginning of the backup.
Using this flag enables this backup to be appended to the previous backup
performed. The tape is not automatically rewound at the end of the backup.
The -n option is always ignored when backing up to a non-tape device.
-p
Indicates that the data should be packed before being written to the media.
Appendix A. Commands
A-31
This typically reduces size of backup between 25% and 40%. For increased
performance, do not use this options when backing up to a device that
provides hardware data compression.
-O
If a filename was specified as the output device and the filename already
exists, overwrites the previous file with the new backup.
-r user=[u|a]
Specifies user read permission when backing up to a disk image file. “u”
indicates only the specified user can read the file. “a” indicates that any user
can read the file.
-r host=[h|a]
Specifies host read permission when backing up to a disk image file. “h”
indicates only this host can read the file. “a” indicates any host can read the
file.
-U “host”
The name of the host to be backed up (such as the backup system).
-v Specifies that file names should be listed on the screen as the files are being
backed up. This flag cannot be used with the -x flag.
-x Specifies that the progress indicator should be shown on the screen, displaying
the approximate size and time to read the entire contents of the media and the
amount completed. This flag cannot be used with the -v flag.
Examples
1. To perform a backup of the /home/tony directory to the backup directory
/usr/lpp/sysback/bf/root, allowing only this user and host to read the file, and
to display a progress indicator, type:
mkdirback -f /usr/lpp/sysback/bf/root -x -r host=h -r user=u -I
tony_tuesday /home/tony
The above will create a file called
/usr/lpp/sysback/bf/root/FD.hostname.tony_tuesday.
2. To perform a remote push backup to virtual device vdev2 on host jupiter and to
make this a compressed backup of the /data/file1 and /data/file2 files, type:
mkdirback -Dph jupiter -f vdev2 /data/file1 /data/file2
3. To perform a backup to rmt1 on the local system of only the files in the /usr
and /var directories that have changed within the last seven days, type:
mkdirback -nm7 -f /dev/rmt1 /usr /var
4. To initiate a pull backup of /home on host lasher, from host sysback1, and send
the data to rmt1 of the host sysback1, type the following from sysback1:
mkdirback -U lasher -f /dev/rmt1 /home
5. To initiate a pull backup of /home on host lasher, from host sysback1, and send
the data to rmt1 of the host shappy, type the following from sysback1:
mkdirback -U lasher -h shappy -f /dev/rmt1 /home
Related Information
The sysrestore, cfgremsvs, cfgremaccess, chexcludelist, and cfgremrootaccess
commands.
A-32
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
mkjfsback
Purpose
Performs a backup of one or more filesystems.
Syntax
mkjfsback
-v
-x
-b
-c
-d
-e
-h
-i
-l
-M
-n
-O
-p
-r
-U
″size″
″termdev″
″description″
-f ″device″
-f ″directory″
-f ″file″
-I ″ID″
″hostname″
″level″
″readAccess″
″host″
mount point
Description
The mkjfsback command is used to perform a backup of one or more filesystems.
The mount point parameter, which is the directory name where each filesystem is
mounted (such as /home), indicates the filesystems to include. You can back up
multiple filesystems by appending each mount point to the end of the command,
separated by spaces.
The backup can be performed to a tape drive, virtual device, or disk image file.
When specifying a disk image file, you can enter the full pathname of the file to
create or overwrite. When doing so, the filename must begin with the prefix “FS”
to indicate the file is a filesystem backup. You can also specify only the backup
directory and a unique ID. When doing so, a filename will be created for you.
Refer to “Backups to Disk Image Files” on page 2-6 for information on the
filename created. If you are creating a file and a file by the same name already
exists, specify the -O option to overwrite the existing file.
The backup can also be performed to a remote device, virtual device, or disk
image file. If Remote Services has been configured on both the local and server
system using the cfgremsvs command, and you have defined a backup device on
the server for use by this system using the cfgremaccess command, then you can
also select a hostname using the -h option. The backup will then be written to the
device or filename on the specified server, provided the server has enabled this
system access to the specified device or directory.
A level 0 backup must be performed before any level 1-9 can be performed. Refer
to “Understanding Incremental Backups” on page 4-1 for additional details on
incremental backups.
Appendix A. Commands
A-33
This command does not back up files, or files within directories, that are specified
in the exclude list. See the chexcludelist command for information on creating
exclude lists.
The backup can also be performed as a “pull” backup to enable central
management of backups. The machine that “pulls” the backup is called the
“initiator,” while the machine that is backed up is called the “backup system.” The
machine that receives the backup data is called the “destination system.”
The initiator is the machine where the mkdirback command is issued. The backup
system is referenced in the command syntax by the -U “host” flag. The data
destination system is referenced by the -h “host flag”.
When only the -U “host” flag is specified, the -f “device” flag refers to a device on
the initiator system. In this case, the backup destination and the initiator are the
same system. This is known as a “two-way pull backup.”
When the -U “host” and -h “host” flags are specified, the -f “device” flag refers to
a device on the host specified by the -h “host” flag. In this case, the backup
destination and the initiator are different systems. This is known as a “three-way
pull backup.”
Remote Services must be configured, using the cfgremsvs command, on all of the
machines involved in the backup operation, the initiator, the backup system, and
the data destination machines. The backup destination machine must have the
Server Options of Remote Services configured to allow the backup system machine
to send its data to devices on the backup destination machine. You can do this
with the Add or Change Client Host Access to this Server option or the
cfgremaccess command.
The backup system machine must have the Client Options of Remote Services
configured to specify the backup destination machine as a backup host. You can do
this with the Add or Change List of Remote Backup Servers option or the
cfgremserver command.
The backup system must also configure Remote Command Access to allow the
initiator machine to pull the backup.
Flags
-b “size”
Specifies the buffer size in K-bytes (1024-bytes). This value either increases or
decreases the amount of data that is written to the output device at one time.
Some devices with small buffers might require you to reduce this value, while
other devices with very large buffers can benefit from increased performance
by increasing this value. If you are unsure, use the default value (64 Kbytes).
-c “termdev”
If the backup spans multiple tape volumes, the volume prompt is sent to the
termdev device (such as /dev/tty0). If the -h hostname option is specified, this
flag specifies the device on the remote host.
-d “description”
A custom description to be included in the backup header. If the description
includes spaces, the entire description must be enclosed in double quotation
marks (″).
A-34
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
-e Forwards the tape to the end of the last Sysback backup. Used when stacking
multiple backup images on a tape.
-f “device, file or directory”
Specifies the output device name, filename, or directory for the backup. The
specified output option can be on the local system or server (if -h option
provided). The device name can be a tape drive or virtual device name. If you
specify a directory, you must also specify an ID using the -I flag.
-g “filename”
Indicates the name of an LVM Information File previously created and
customized for this backup. Refer to the mkvginfo and editlvminfo
commands.
-h “hostname”
Name of the remote backup server host to receive the backup data (such as the
data destination host). This is the host that provides the backup device
specified by the -fflag.
-i
Indicates to dynamically create a backup file ID when writing backups to disk
image file. This flag supersedes the -I“ID” flag.
-I “ID”
This option is used only if a directory is specified as the output device. The ID
is included in the filename created to make it unique from other disk image
file backups. If a backup in the same directory with this ID already exists,
specify the -O flag to overwrite the previous backup.
-l “level”
Specifies the backup level (default is 0, or all files). Valid values are 0-9. Only
files created or changed since the last (level - 1) backup was performed.
-M
Retains physical partition mapping, by default, when the logical volume is
recreated from this backup. Unless the logical volume was originally created
using a physical partition map, it is generally recommended that it not be
recreated using the same partitions, as this would preserve fragmentation that
develops over time as logical volumes and filesystems are expanded. You can
specify whether or not partition mapping will be used before recreating the
logical volume.
-n
Indicates that the tape should not be rewound at the beginning of the backup.
This enables the backup to be appended to the previous backup performed.
The tape is not automatically rewound at the end of the backup. The -n option
is always ignored when backing up to a non-tape device.
-O
Overwrites the previous file, if a filename was specified as the output device,
and the filename already exists.
-p
Indicates that the data should be packed before being written to the media.
This typically reduces size of backup between 25% and 40%. For increased
performance, do not use this option when backing up to a device that provides
hardware data compression.
-r user=[u|a]
Specifies user read permission when backing up to a disk image file. “u”
indicates only the specified user can read the file. “a” indicates that any user
can read the file.
Appendix A. Commands
A-35
-r host=[h|a]
Specifies host read permission when backing up to a disk image file. “h”
indicates only the specified host can read the file. “a” indicates that any host
can read the file.
-U “host”
Name of the host to be backed up (such as the backup system).
-v Specifies that file names should be listed on the screen as the files are being
backed up. This flag cannot be used with the -x flag.
-x Specifies that the progress indicator should be shown on the screen, displaying
the approximate size and time to read the entire contents of the media and the
amount completed. This flag cannot be used with the -v flag.
Examples
1. To perform a local backup of the /home filesystem to the directory
/usr/lpp/sysback/bf/mars, indicating that only this host and user can read the
file and to display a progress indicator, type:
mkjfsback -f /usr/lpp/sysback/bf/mars -I home4me -x \
-r host=h -r user=u /home
The filename created will be /usr/lpp/sysback/bf/mars/FS. hostname.home4me.
2. To perform a level 1 backup of the /data/files and /sales filesystems as a
remote push backup to /dev/rmt1 on the host jupiter, type:
mkjfsback -l1 -h jupiter -f rmt1 /data/files /sales
Note that a level 0 backup must have previously been created before a level 1
can be performed.
3. To perform a level 9 backup of the /home filesystem to the local virtual device
vdev1, type:
mkjfsback -nl9 -f vdev1 /home
Note that the -n option is specified. This prevents the tape drive from being
rewound before the backup begins, enabling this backup to be “stacked” on the
end of the previous backup image.
4. To initiate a pull backup of the /home filesystem on host lasher, from host
sysback1, and send the data to rmt1 of the host sysback1, type the following from
sysback1:
mkjfsback -U lasher -f /dev/rmt1 /home
5. To initiate a pull backup of the /home filesystem on host lasher, from host
sysback1, and send the datat to rmt1 of the host shappy, type the following
command from sysback1:
mkjfsback -U lasher -h shappy -f /dev/rmt1 /home
Related Information
The sysrestore, cfgremsvs, cfgremaccess, mkvginfo, editlvminfo,cfgremrootaccess,
and chexcludelist commands.
A-36
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
mklvback
Purpose
Performs a backup of logical volumes.
Syntax
mklvback
-b
-c
-d
-e
-g
-h
-i
-M
-n
-O
-p
-r
-U
-v
″size″
″termdev″
″description″
-f ″device″
-f ″directory″
-f ″file″
LVname
-I ″ID″
″filename″
″hostname″
″readAccess″
″host″
Description
The mklvback command is used to back up data from one or more logical
volumes. This option reads the entire logical volume, copying all “raw” data to the
specified output device or file. Many database applications use a raw logical
volume to store data, but AIX is unaware of the format of the data within the
logical volume. Therefore, this option copies the entire contents of the logical
volume, even though the application might have written to only a small portion of
the total logical volume size.
The backup can be performed to a tape drive, virtual device, or disk image file.
When specifying a disk image file, you can enter the full pathname of the file to
create or overwrite. When doing so, the filename must begin with the prefix “LV”.
to indicate the file is a logical volume backup. You can also specify only the
backup directory and a unique ID. When doing so, a filename will be created for
you. Refer to “Backups to Disk Image Files” on page 2-6 for information on the
filename created. If you are creating a file and a file by the same name already
exists, specify the -O option to overwrite the existing file.
The backup can also be performed to a remote device, virtual device, or disk
image file. If Remote Services has been configured on both the local and server
system using the cfgremsvs command, and you have defined a backup device on
the server for use by this system using the cfgremaccess command, then you can
also select a hostname using the -h option. The backup will then be written to the
device or filename on the specified server, provided the server has enabled this
system access to the specified device or directory.
You must supply one or more logical volume names with this command. Do not
include the “/dev/” prefix with the logical volume name. If specifying more than
one logical volume, append each logical volume to the end of the command,
separated by spaces.
Appendix A. Commands
A-37
The backup can also be performed as a “pull” backup to enable central
management of backups. The machine that “pulls” the backup is called the
“initiator,” while the machine that is backed up is called the “backup system.” The
machine that receives the backup data is called the “destination system.”
The initiator is the machine where the mkdirback command is issued. The backup
system is referenced in the command syntax by the -U “host” flag. The data
destination system is referenced by the -h “host flag”.
When only the -U “host” flag is specified, the -f “device” flag refers to a device on
the initiator system. In this case, the backup destination and the initiator are the
same system. This is known as a “two-way pull backup.”
When the -U “host” and -h “host” flags are specified, the -f “device” flag refers to
a device on the host specified by the -h “host” flag. In this case, the backup
destination and the initiator are different systems. This is known as a “three-way
pull backup.”
Remote Services must be configured, using the cfgremsvs command, on all of the
machines involved in the backup operation, the initiator, the backup system, and
the data destination machines. The backup destination machine must have the
Server Options of Remote Services configured to allow the backup system machine
to send its data to devices on the backup destination machine. You can do this
with the Add or Change Client Host Access to this Server option or the
cfgremaccess command.
The backup system machine must have the Client Options of Remote Services
configured to specify the backup destination machine as a backup host. You can do
this with the Add or Change List of Remote Backup Servers option or the
cfgremserver command.
The backup system must also configure Remote Command Access to allow the
initiator machine to pull the backup.
Flags
-b “size”
Specifies the buffer size in K-bytes (1024-bytes). This value either increases or
decreases the amount of data that is written to the output device at one time.
Some devices with small buffers might require you to reduce this value, while
other devices with very large buffers can benefit from increased performance
by increasing this value. If you are unsure, use the default value (64 Kbytes).
-c “termdev”
If the backup spans multiple tape volumes, the volume prompt is sent to the
termdev device (such as /dev/tty0). If the -h hostname option is specified, this
flag specifies the device on the remote host.
-d “description”
A custom description to be included in the backup header. If the description
includes spaces, the entire description must be enclosed in double quotation
marks (″).
-e Forwards to the end of the last SysBack backup on the tape. Used when
stacking multiple backups on one tape.
-f “device, file or directory”
Specifies the output device name, filename, or directory for the backup. The
A-38
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
specified output option can be on the local system or server (if -h option
provided). The device name can be a tape drive or virtual device name. If you
specify a directory, you must also specify an ID using the -I flag.
-g “filename”
Indicates the name of an LVM Information File previously created and
customized for this backup. Refer to the mkvginfo and editlvminfo
commands.
-h “hostname”
Name of the remote backup server host to receive the backup data (such as the
data destination host). This is the host that provides the backup device
specified by the -f flag.
-i
Indicates to dynamically create a backup file ID when writing backups to disk
image file. This flag supersedes the -I“ID” flag.
-I “ID”
This option is used only if a directory is specified as the output device. The ID
is included in the filename created to make it unique from other disk image
file backups. If a backup in the same directory with this ID already exists,
specify the -O flag to overwrite the previous backup.
-l “level”
Specifies the backup level (default is 0, or all files). Valid values are 0-9. Only
files created or changed since the last (level - 1) backup was performed.
-n
Indicates that the tape should not be rewound at the beginning of the backup.
This enables the backup to be appended to the previous backup performed.
The tape is not automatically rewound at the end of the backup. The -n option
is always ignored when backing up to a non-tape device.
-O
Overwrites the previous file, if a filename was specified as the output device,
and the filename already exists.
-p
Indicates that the data should be packed before being written to the media.
This typically reduces size of backup between 25% and 40%. For increased
performance, do not use this option when backing up to a device that provides
hardware data compression.
-r user=[u|a]
Specifies user read permission when backing up to a disk image file. “u”
indicates only the specified user can read the file. “a” indicates that any user
can read the file.
-r host=[h|a]
Specifies host read permission when backing up to a disk image file. “h”
indicates only the specified host can read the file. “a” indicates that any host
can read the file.
-U “host”
Name of the host to be backed up (such as the backup system).
-x Specifies that the progress indicator should be shown on the screen, displaying
the approximate size and time to read the entire contents of the media and the
amount completed. This flag cannot be used with the -v flag.
-M
Retains physical partition mapping, by default, when the logical volume is
Appendix A. Commands
A-39
recreated from this backup. Unless the logical volume was originally created
using a physical partition map, it is generally recommended that it not be
recreated using the same partitions, as this would preserve fragmentation that
develops over time as logical volumes and filesystems are expanded. You can
specify whether or not partition mapping will be used before recreating the
logical volume.
Examples
1. To perform a local backup of the lv02 logical volume to the directory
/usr/lpp/sysback/bf/root, indicating that only this user and host can read the
file and displaying a progress indicator, type:
mklvback -f /usr/lpp/sysback/bf/root -I lv02 -x -r host=h -r user=u lv02
The filename created will be /usr/lpp/sysback/bf/root/LV.hostname.lv02.
2. To perform a remote push backup to the virtual device vdev2 of the host jupiter
and to make this a compressed backup of the logical volumes datalv and
datalv2, type:
mklvback -nph jupiter -f vdev2 datalv datalv2
Note that the -n flag is specified. This prevents the tape drive from being
rewound before the backup begins, enabling this backup to be “stacked” on the
end of the previous backup image.
3. To initiate a pull backup of the /home filesystem on host lasher, from host
sysback1, and send the data to rmt1 of the host sysback1, type the following from
sysback1:
mklvback -U lasher -f /dev/rmt1 /home
4. To initiate a pull backup of the /home filesystem on host lasher, from host
sysback1, and send the data to rmt1 of the host shappy, type the following
command from sysback1:
mklvback -U lasher -h shappy -f /dev/rmt1 /home
Related Information
The sysrestore, cfgremsvs, cfgremaccess, mkvginfo, cfgremrootaccess, and
editlvminfo commands.
A-40
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
mksbnetboot
Purpose
Creates or updates boot images on the server for booting network clients.
Configures the server to enable network booting of client machines.
Syntax
mksbnetboot -B -h ″hostname″
-a
-d
-g
-s
-u
-S
-T
-k
-n
-I
-D
″address″
″net device″
″gateway″
″subnet″
″server addr″
″platform″
″kernel″
″NIM SPOT″
″NIM LPPSOURCE″
Description
The mksnetboot command configures network boot clients and the boot images
required to boot them. Both NIM Resource and Classic methods are handled. See
Chapter 13, “Network Boot/Installation Configuration”, on page 13-1 for an
explanation of NIM Resource and Classic netboot methods.
A boot image is created when a new client is configured. Boot images might need
to be updated if any server software is updated. For example, if the NIM spot has
a new software level applied to it, the corresponding boot images built from that
spot should be updated.
For the Classic netboot method, the boot images should be updated if any software
is applied to the netboot server (its /usr filesystem). The default case is to update
all known boot images if an update is requested.
Configuring a Network Boot Client Machine
Classic Method: Specifying the -h, -a and -d arguments generates the information
needed to respond to the client’s BOOTP request and assigns the client the specific
boot image used to boot the client, depending on the client’s machine type.
However, if you do not specify the -a option, the command requires, at a
minimum, the -S, and optionally the -s, and -g flags to add a client using this
method.
You only need to run this command for clients who will boot from the server. For
clients that will boot from other media, such as installation diskettes or tape but
will only install from this server, you do not need to run this command.
If the network adapter has been changed in the client system, you might need to
update the server to reflect the new hardware address of the client machine. You
can update the boot client information if already configured by following the same
steps as for a new client. This overwrites the previous information.
Appendix A. Commands
A-41
NIM Method: When you add a network boot client that uses NIM resources, the
-d, -T, -k, -S, -s, -g, and -a flags are not valid. The information that would be
contained in these flags is contained in the NIM resource definition for the client.
The required flags to add a NIM Resource Network Boot client are -h, -n, and -l.
Rebuilding Existing Network Boot Images
If you update or install software on the boot source, you must also update the boot
image. For the Classic boot method, the boot source is the /usr directory of the
boot server. For the NIM Resource Network Boot, the boot source is the NIM
SPOT resource.
To rebuild the boot image for either method, use one of the following commands.
To rebuild all existing boot images, type:
mksbnetboot -B ALL
To rebuild a specific build image, type:
mksbnetboot -B <image name>
A Note about the Network Hardware Address
The -a flag is used to specify the adapter “hardware” address, which differs for
each adapter manufactured. The adapter address is needed by the server before it
can respond to BOOTP requests. If you do not know the client network adapter
hardware address, follow the instructions in “Determining the Network Adapter
Hardware Address” on page B-5.
Enter this parameter only for PCI-based RISC System (rspc) clients, or if you want
the client to boot using a broadcast BOOTP request. By entering the adapter
hardware address for a non-rspc client, that client can be booted without filling out
the server information in the BOOTP screen at system startup. This is
accomplished by sending a broadcast BOOTP request that the server responds to
by identifying the hardware address of the client’s network adapter. This
information is not used by non-RISC PC clients requiring a gateway to reach the
server.
Flags
-a “address”
Specifies the adapter hardware address of your network adapter. Note that this
is not referring to the Internet (IP) address, but the hardware adapter address.
This information is required on PCI-based RISC System (rspc) clients and is
not used by other clients requiring a gateway to reach the server.
-d “net device”
Specifies the name of the network adapter for the network the clients will be
booted from. You can also specify a network adapter type of either ethernet or
token-ring if a device of the client’s network adapter type does not exist on the
server.
-g “gateway”
Gateway address the client must use to reach this server, if any.
-h “hostname”
Specifies the name of the host to which the boot image is to be assigned or the
client to be removed from the network boot configuration.
-k “kernel”
Indicates the kernel type, either “up” for uniprocessors or “mp” for
multiprocessors.
A-42
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
-u
Specifies that the hostname is to be unconfigured as a network boot client.
-s “subnet”
Subnet mask of the client. The default is 255.255.255.0.
-B
Specifies to update a network boot image.
-n
The name of the NIM spot. The –n flag implies NIM netboot mode and also
requires the –n flag. Valid spot names can be discovered using the command
lsnim –t spot.
-l
The name of the NIM lpp_source. The –l flag also requires the –n flag. Use the
command lsnim –t spot to find valid lpp_source names.
-S “server addr”
IP address of the server. The default is the primary IP address of the current
system. Specify this value only if the server has multiple IP addresses. Specify
the address the client will use to contact the server.
-T “platform”
The platform type of the client. Use the -k kernel flag to indicate the kernel
type.
The following platform types are currently supported:
rs6k
Classic RISC System/6000
rspc
PCI-based (PC) RISC System/6000
chrp
Common Hardware Reference Platform
-D
Indicates to build the network boot image with the AIX Kernel Debugger
enabled. Use this flag for diagnostic purposes only. To use this flag, you must
have a terminal attached to the machine’s s1 port in order to interact with the
kernel debugger to initiate the network boot. It is important to note the
diagnostic value that is provided at the end of the boot image build. This
alphanumeric word is a storage address used by the debugger at boot time.
Examples
1. To add a network boot client named “mars” and create a network boot image
for a Multiprocessor RISC System/6000 platform on a token ring network using
the Classic netboot method, type:
mksbnetboot -h mars -T rspc -K up -d token ring -S 9.19.134.93
2. To add a network boot client named “lasher” using the NIM resource method
with a NIM SPOT called “spot_aix433” and an LPPSOURCE called
“lpp_aix433”, type:
mksbnetboot -h lasher -n spot_aix433 -l lpp_aix433
3. To rebuild all previously configured network boot images, type:
mksbnetboot -B all
Files
/etc/bootptab
Contains entries for response to BOOTP requests
received from the network.
/tftpboot/hostname
A symbolic link to an actual network boot image
file for the specific hostname.
Appendix A. Commands
A-43
/usr/lpp/sysback/netinst/boot
A-44
Directory where the actual boot images are stored.
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
mkvgback
Purpose
Performs a backup of one or more volume groups.
Syntax
mkvgback
-v
-x
-b
-c
-d
-e
-g
-h
-i
-l
-m
-M
-n
-O
-p
-r
-U
-X
″size″
″termdev″
″description″
-f ″device″
-f ″directory″
-f ″file″
-I ″ID″
″filename″
″hostname″
″level″
″days″
″readAccess″
″host″
VGname
Description
The mkvgback command is used to find and back up all files belonging to the
specified volume group or groups.
The backup can be performed to a tape drive, virtual device, or disk image file.
When specifying a disk image file, you can enter the full pathname of the file to
create or overwrite. When doing so, the filename must begin with the prefix “VG”.
to indicate the file is a volume group backup. You can also specify only the backup
directory and a unique ID. When doing so, a filename will be created for you. Refer
to “Backups to Disk Image Files” on page 2-6 for information on the filename
created. If you are creating a file and a file by the same name already exists,
specify the -O option to overwrite the existing file.
The backup can also be performed to a remote device, virtual device, or disk
image file. If Remote Services has been configured on both the local and server
system using the cfgremsvs command, and you have defined a backup device on
the server for use by this system using the cfgremaccess command, then you can
also select a hostname using the -h option. The backup will then be written to the
device or filename on the specified server, provided the server has enabled this
system access to the specified device or directory.
The -l level option indicates that an incremental backup will be performed.
A level 0 backup must be performed before any level 1-9 can be performed. Refer
to “Understanding Incremental Backups” on page 4-1 for additional details on
incremental backups.
Appendix A. Commands
A-45
Files, or files within directories, that are specified in the exclude list will not be
backed up with this command. See the chexcludelist command for information on
creating exclude lists.
The backup can also be performed as a “pull” backup to enable central
management of backups. The machine that “pulls” the backup is called the
“initiator,” while the machine that is backed up is called the “backup system.” The
machine that receives the backup data is called the “destination system.”
The initiator is the machine where the mkdirback command is issued. The backup
system is referenced in the command syntax by the -U “host” flag. The data
destination system is referenced by the -h “host flag.”
When only the -U “host” flag is specified, the -f “device” flag refers to a device on
the initiator system. In this case, the backup destination and the initiator are the
same system. This is known as a “two-way pull backup.”
When the -U “host” and -h “host” flags are specified, the -f “device” flag refers to
a device on the host specified by the -h “host” flag. In this case, the backup
destination and the initiator are different systems. This is known as a “three-way
pull backup.”
Remote Services must be configured, using the cfgremsvs command, on all of the
machines involved in the backup operation, the initiator, the backup system, and
the data destination machines. The backup destination machine must have the
Server Options of Remote Services configured to allow the backup system machine
to send its data to devices on the backup destination machine. You can do this
with the Add or Change Client Host Access to this Server option or the
cfgremaccess command.
The backup system machine must have the Client Options of Remote Services
configured to specify the backup destination machine as a backup host. You can do
this with the Add or Change List of Remote Backup Servers option or the
cfgremserver command.
The backup system must also configure Remote Command Access to allow the
initiator machine to pull the backup.
Flags
-b “size”
Specifies the buffer size in K-bytes (1024-bytes). This value either increases or
decreases the amount of data that is written to the output device at one time.
Some devices with small buffers might require you to reduce this value, while
other devices with very large buffers can benefit from increased performance
by increasing this value. If you are unsure, use the default value (64 Kbytes).
-c “termdev”
If the backup spans multiple tape volumes, the volume prompt is sent to the
termdev device (such as /dev/tty0). If the -h hostname option is specified, this
flag specifies the device on the remote host.
-d “description”
A custom description to be included in the backup header. If the description
includes spaces, the entire description must be enclosed in double quotation
marks (″).
A-46
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
-e Forwards the tape to the end of the last Sysback backup. Used when stacking
multiple backup images on a tape.
-f “device, file or directory”
Specifies the output device name, filename, or directory for the backup. The
specified output option can be on the local system or server (if -h option
provided). The device name can be a tape drive or virtual device name. If you
specify a directory, you must also specify an ID using the -I flag.
-g “filename”
Indicates the name of an LVM Information File previously created and
customized for this backup. Refer to the mkvginfo and editlvminfo
commands.
-h “hostname”
Name of the remote backup server host to receive the backup data (such as the
data destination host). This is the host that provides the backup device
specified by the -f flag.
-i
Indicates to dynamically create a backup file ID when writing backups to disk
image file. This flag supersedes the -I“ID” flag.
-I “ID”
This option is used only if a directory is specified as the output device. The ID
is included in the filename created to make it unique from other disk image
file backups. If a backup in the same directory with this ID already exists,
specify the -O flag to overwrite the previous backup.
-l “level”
Specifies the backup level (default is 0, or all files). Valid values are 0-9. Only
files created or changed since the last (level - 1) backup was performed.
-M
Retains physical partition mapping, by default, when the logical volume is
recreated from this backup. Unless the logical volume was originally created
using a physical partition map, it is generally recommended that it not be
recreated using the same partitions, as this would preserve fragmentation that
develops over time as logical volumes and filesystems are expanded. You can
specify whether or not partition mapping will be used before recreating the
logical volume.
-n
Indicates that the tape should not be rewound at the beginning of the backup.
This enables the backup to be appended to the previous backup performed.
The tape is not automatically rewound at the end of the backup. The -n option
is always ignored when backing up to a non-tape device.
-O
Overwrites the previous file, if a filename was specified as the output device,
and the filename already exists.
-p
Indicates that the data should be packed before being written to the media.
This typically reduces size of backup between 25% and 40%. For increased
performance, do not use this option when backing up to a device that provides
hardware data compression.
-r user=[u|a]
Specifies user read permission when backing up to a disk image file. “u”
indicates only the specified user can read the file. “a” indicates that any user
can read the file.
Appendix A. Commands
A-47
-r host=[h|a]
Specifies host read permission when backing up to a disk image file. “h”
indicates only the specified host can read the file. “a” indicates that any host
can read the file.
-U “host”
Name of the host to be backed up (such as the backup system).
-v Specifies that file names should be listed on the screen as the files are being
backed up. This flag cannot be used with the -x flag.
-x Specifies that the progress indicator should be shown on the screen, displaying
the approximate size and time to read the entire contents of the media and the
amount completed. This flag cannot be used with the -v flag.
-X
Excludes all non-JFS (journaled file system) logical volumes, or “raw” logical
volumes, from the backup.
Examples
1. To perform a level 1 backup of the vg00 volume group to the directory
/usr/lpp/sysback/bf/saturn/root, listing the files as they are backed up, type:
mkvgback -vf /usr/lpp/sysback/bf/saturn/root -I vg00_lev1 -l1 vg00
The file created will be
/usr/lpp/sysback/bf/saturn/root/VG.hostname.vg00_lev1.
2. To perform a remote backup to the virtual device vdev3 on the host jupiter that
is a compressed backup of the uservg1 and uservg2 volume groups and to
display a progress indicator, type:
mkvgback -pxh jupiter -f vdev3 uservg1 uservg2
3. To perform a remote backup to the virtual device /dev/rmt1 of the data1 volume
group, including only the files created or changed in the last two days, type:
mkvgback -nf /dev/rmt1 m2 data1
4. To initiate a pull backup of the /home filesystem on host lasher, from host
sysback1, and send the data to rmt1 of the host sysback1, type the following from
sysback1:
mkvgback -U lasher -f /dev/rmt1 /home
5. To initiate a pull backup of the /home filesystem on host lasher, from host
sysback1, and send the datat to rmt1 of the host shappy, type the following
command from sysback1:
mkvgback -U lasher -h shappy -f /dev/rmt1 /home
Related Information
The sysrestore, cfgremsvs, cfgremaccess, mkvginfo, cfgremrootaccess, editlvminfo
and chexcludelist commands.
A-48
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
mkvginfo
Purpose
Generates information needed in recreating volume groups and logical volumes
from backups.
Syntax
mkvginfo
-v VGname -l LVname
-f
-M
-o ″filename″
Description
The mkvginfo command is automatically executed by each SysBack backup
command that stores LVM information on the backup. This information contains
the volume group, logical volume, filesytem, and physical volume attributes that
can be later used to recreate any of the preceding from the backup media.
The arguments supplied to the command and the resulting output depend on the
type of backup performed. This command sends the results to standard output
(stdout) by default, unless the -o filename option is specified.
This command generates LVM information for either a list of volume groups, if the
-v VGnames option is specified, or for a list of logical volumes if the -l LVnames
option is specified. If the logical volumes contain filesystems, the filesystem
attributes are recorded as well.
The output contains a label, indicating the record type, followed by the attributes
for the particular label. The contents of the /etc/filesystems file, for the logical
volumes involved, are appended to the end of the file.
The information contained in the output is as follows:
System Information (one of each of the following):
TYPE
Backup type (VG, LV, SYSTEM, POWER)
VERSION
Version and Release of AIX (such as “4.1”)
PLATFORM
Platform Type (such as rs6k)
KERNEL
Kernel Type (up/mp)
SBVER
SysBack Version, Release and Modification Level
(such as 040101)
Table of contents (one record, indicates volume group data, if any, included on
backup):
TOC
Static Header
vgname(s)
List of volume groups (separated by spaces)
x
Static Trailer
Physical Volumes (one record for each):
PV
Static Header
PVid
Physical volume ID
Physical Volumes (one record for each):
Appendix A. Commands
A-49
Location
Hdiskname
Location Code (address)
Hdisk Name (not used)
Volume Groups (one record for each):
VG
Static Header
vgname
Volume Group name
autoon
Auto varyon at system startup? (Y/N)
ppsize
Physical partition (in MB)
vgopt
[C]reate/[I]mport/[X]Ignore
quorum
Are quoroms active? (Y/N)
concurrent
Concurrent-capable? (y/n)
autoconc
Auto-Concurrent Varyon? (y/n)
pvlist
List of physical volume IDs
Logical Volumes (one record for each)
LV
Static Header
priority
(01=mapped, 02=striped, 03=boot, 04=jfslog,
10=paging, 20=jfs, 90=dump)
mountpt
Filesystem mount point or ’-’ if not jfs
vgname
Volume group name
lvname
Logical volume name
type
Logical volume type (see priority)
createopt
Create this LV? (Y/N)
origmb
Original size in MB
lps
Size to create in LPs (can change by user)
copies
Number of copies (#PPs per LP)
minips
Minimum LPs to contain filesystem data (1 if not
jfs)
intra
Intra-pv allocation policy (ie/im/c/om/m/oe/e)
inter
Inter-pv allocation policy (m/x)
upper
Upperbound - Max PVs if maximum inter policy
mzxlps
Maximum LPs for this LV
mwc
Mirror-write-consistency on? (y/n)
bbpolicy
Bad block relocation? (y/n)
relocatable
Can LV be relocated? (y/n)
strict
Can multiple copies be place on one PV? (y/n)
verify
Write verify? (y/n)
stripesz
Stripe Size
stripewd
Stripe width
ppmap
Create using physical partition map? (y/n)
LVpvlist
Physical volume IDs
Filesystems (JFS only - one record for each) – AIX Version 4 only:
FS
Static Header
mountpt
Mount point (same as in LV record)
vgname
Volume Group name
lvname
Logical Volume name
fragsize
Fragment Size
nbpi
Number of bytes per inode
compress
Compress? (LZ, no)
bfsupport
Big File Support? (y/n) (AIX 4.2+)
Filesystems (JFS only - one record for each) – AIX Version 4 only:
agsize
Allocation Group Size (#MB) (AIX 4.2+)
A-50
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Physical Partition Maps (one or more each logical volume – each not to exceed
50 PPs):
PP
Static Header
lvname
Logical Volume name
ppno:ppvid
Physical partition number: physical volume ID (one
for each PP in the LV)
Filesystem stanzas (one record for each):
Stanza taken directly from /etc/filesystems file
The information produced by this command is placed on the beginning of each
backup and can be read from the backup using the sbread command with the -T
flag.
Including a customized LVM information file on a backup: You can create and
customize an LVM information file to be included on a SysBack backup. This
information enables you to later recreate volume groups, logical volumes, or
filesystems using the customized attributes without requiring the user to make the
changes manually using the menu options.
To do so, use the following steps:
1. Use the mkvginfo command to create an LVM information file. Use the -o
filename option to specify an output filename.
2. Use the editlvminfo command to edit the file using the menus and options
described in Chapter 18, “Changing the Volume Group, Logical Volume and
Filesystem Attributes”, on page 18-1.
3. Create the backup using the mkjfsback, mklvback, mkvgback or sysback
commands. Use the -g filename option with the command to include the
customized file instead of generating a new one.
Flags
-f
Indicates the command should continue even if filesystems are not mounted.
Ordinarily, this command fails if a filesystem is not mounted.
-l
Indicates that a list of logical volumes is supplied and that the LVM
information should be generated only for the specified logical volumes and
filesystems, if applicable.
-M
Indicates that the default when recreating the logical volume or logical
volumes is to recreate using physical partition maps. If not specified, the
default is to recreate the logical volumes using contiguous partitions on the
physical volumes, whenever possible.
-o “filename”
The name of the file where the output should be sent. If not specified, output
is sent to standard output.
-v Indicates that a list of volume groups is supplied and that the LVM
information should be generated for the specified volume groups.
Files
/etc/filesystems
Contains filesystem stanzas that are copied to the
output of this command for each logical volume
containing a filesystem.
Appendix A. Commands
A-51
Related Information
The mkjfsback, mklvback, mkvgback, sysback, editlvminfo and sbread
commands.
A-52
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
readsbheader
Purpose
Reads the header label on any SysBack backup.
Syntax
readsbheader
-c ″termdev″
-d
-h ″hostname″
-i ″sequence″
-l
-n
-q
-f ″device″
Description
The readsbheader command can be used to read the header information on any
SysBack backup. It can be used to list the general header information, the list of
images (logical volumes and filesystems) included on the backup, and the list of
files and directories included on the backup. This command is useful for reading
the contents of an unknown backup tape when no printed label was affixed.
If the backup media is a tape that contains multiple backups, you can read the
header information from any backup on the tape by using the -i “sequence”
option. The sequence refers to the backup sequence number, “1” being the first
backup on the tape and so on.
Flags
-c “termdev”
If the backup contains multiple tape volumes, the volume prompt is sent to the
display indicated by “termdev” (such as /dev/tty0 or /dev/lft0). If the -h
“hostname” option is specified, this flag specifies the display on the remote
host.
-d
Indicates that a list of logical volumes and filesystems included on the backup
should be displayed also.
-f “device or file”
Tape device, virtual device, or filename containing the backup. If a filename is
specified, the information is read from the file on the disk with the “.TOC”
extension. You do not need to supply this extension.
-h “hostname”
Hostname of backup server to which the remote backup device is attached.
-i “sequence”
Indicates the backup sequence number to read. “1” refers to the first backup on
the tape, and so on. The default is “1.” This option can be specified only when
reading from a tape drive or virtual device.
-l
Indicates that a list of files and directories included on the backup should be
displayed. This file list can be very long. The directories are listed first,
followed by the regular files.
Appendix A. Commands
A-53
-n
Indicates that tape drives should not be rewound prior to reading. This
assumes that the tape drive is already positioned at the beginning of the
backup to read.
-q
Indicates that no status messages should be displayed while searching and
reading the backup media.
Examples
1. To show the general header information on a backup in tape device /dev/rmt0,
on host mars, type:
readsbheader -f rmt0 -h mars
2. To read the contents of the third backup on virtual device vdev2, including the
list of logical volumes, filesystems, directories and files on the backup, type:
readsbheader -dli3 -f vdev2
Related Information
The sbread and sbwrite commands.
A-54
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
remakevg
Purpose
Recreates one or more specified volume groups, logical volumes, or filesystems
from LVM information contained on the backup media.
Syntax
remakevg
-c
-E
-n
-h
-i
-L
-q
-V
″termdev″
-f ″device″
-v VGname -l LVname
″hostname″
″sequence″
″LVname″
″VGname″
Description
Use the remakevg command to recreate one or more volume groups, logical
volumes, or filesystems using information contained on any backup type except for
the file/directory backup. A volume group, logical volume, or filesystem can only
be recreated from a backup containing the specified data, unless the backup is a
system backup, which contains the information needed to recreate any volume
group, logical volume, or filesystem regardless of the data included on the backup.
After the LVM information is read from the backup, a verification ensures that the
correct physical volumes and space are available to create the volume group and
logical volumes. If not, you will be required to change the LVM information to
satisfy the space requirements.
If the verification fails, or if the -E flag is specified, a user-interface enables you to
view or change the LVM information before the volume groups, logical volumes,
or filesystems are created. Refer to Chapter 18, “Changing the Volume Group,
Logical Volume and Filesystem Attributes”, on page 18-1 for details on changing
the LVM information.
If the -E option is not specified, and there are no problems detected when verifying
the consistency between the LVM information and the current system
configuration, the volume groups, logical volumes, and filesystems are recreated
without additional user-interaction. However, unless the -q option is provided, you
will be asked to confirm before any recreation takes place.
Note that this command recreates only the specified volume groups, logical
volumes, and filesystems (if applicable). It does not restore the data from the
backup media. After creating the desired volume group, logical volume, or
filesystem, you can restore the data with the sysrestore command.
Flags
-c “termdev”
If the backup contains multiple tape volumes, the volume prompt is sent to the
Appendix A. Commands
A-55
display indicated by “termdev” (such as /dev/tty0 or /dev/lft0). If the -h
“hostname” option is specified, this flag indicates the display on the remote
host.
-E
Displays the user-interface menus to enable the LVM information to be
changed before the volume groups, logical volumes, or filesystems are created.
-f “device or file”
Tape device, virtual device, or filename containing the backup. If a filename is
specified, the information is read from the file on the disk with the “.TOC”
extension. You need not supply this extension of the filename.
-h “hostname”
Hostname of backup server to which the remote backup device is attached.
-i “sequence”
Indicates the backup sequence number to read. “1” refers to the first backup on
the tape, and so on. The default is “1”. This option can be specified only when
reading from a tape drive or virtual device.
-l
Indicates that logical volumes (and filesystems, if applicable) should be created
and that the logical volume list is provided at the end of the command.
-L “LVname”
To recreate a logical volume under a different name, specify the new LVname.
This enables you to recreate the logical volume even if the original still exists
on the system. This flag is applicable only when recreating a logical volume,
and you can only specify one logical volume name to create when using this
flag.
-v Indicates that volume groups should be created and that a list of volume
groups is provided at the end of the command.
-V “VGname”
If creating a single volume group: To recreate the volume group under a
different name, specify the new VGname. This enables the volume group to be
created even though the original still exists.
If creating a logical volume: Indicates the volume group name the logical
volume will be created in. Used to move a logical volume from one volume
group to another.
-q
Specifies that no status messages should be displayed while reading the
information from the backup media.
-n
Specifies that the tape is already positioned at the beginning of the select
volume group (on stacked tapes).
Examples
1. To recreate a volume group from a backup on tape device /dev/rmt0,
previously called uservg, under the new name newuservg:
remakevg -f /dev/rmt0 -v -V newuservg uservg
2. To recreate the lv00 and lv01 logical volumes from a volume group backup
image file on host venus, placing the logical volume in volume group vg01,
type:
remakevg -l -h venus -V vg01 -f \
/usr/lpp/sysback/bf/VG.saturn.12300200 lv00 lv01
A-56
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
3. To recreate the datavg1 and datavg2 volume groups from the backup on virtual
device vdev0, enabling the user to change the LVM information prior to the
creation, type:
remakevg -Ev -f vdev0 datavg1 datavg2
Related Information
The sysrestore command.
Chapter 18, “Changing the Volume Group, Logical Volume and Filesystem
Attributes”, on page 18-1.
Appendix A. Commands
A-57
sbalog
Purpose
Changes, lists, or displays the SysBack log.
Syntax
sbalog
-o
-l
-C
-f ″path and filename″
-s ″size″
Description
The SysBack log file provides a mechanism to track the completion status of
SysBack backup, restore, and verify operations. For information about the SysBack
log file, see Chapter 11, “Activity Logging”, on page 11-1.
Flags
-o Lists the contents of the SysBack log file.
-l
Displays the current size and location of the SysBack log file.
-C
Changes the size or location of the SysBack log. You must use either the -f or
-s flag with this flag.
-f “path and filename”
Specifies the location of the SysBack log. Used with the -C flag.
-s “size”
Specifies the size of the SysBack log in multiples of 4096 Kbytes.
Examples
1. To change the SysBack log location to /tmp/logs/sysback.log, type:
sbalog -C -f /tmp/logs/sysback.log
2. To change the size of the log to 8192 Kbytes, type:
sbalog -C -s 8192
3. To change the size and location of the log, type:
sbalog -C -f/tmp/logs/sysback.log -s 8192
4. To list the contents of the log, type:
sbalog -o
Files
/var/adm/ras/sysback.log
The default SysBack log file.
Related Information
The cfgremsvs, cfgremaccess, sbread, sbwrite, and cat commands.
A-58
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
sbclient
Purpose
Performs all SysBack remote command execution.
Syntax
sbclient -i
hostname
IPaddress
command-argument
Description
The sbclient command is used to execute all remote commands in SysBack. It can
be used only after Remote Services has been configured on the local and remote
system using the cfgremsvs command, and the remote system has granted
permission to the local user and host to perform remote commands using the
cfgremaccess command.
When cfgremsvs command is run, a user called sbnet is created. All remote
commands are executed under this user. There is no need for the username on the
local system to exist on the remote system for the local user to execute remote
commands. The sbnet user must have UNIX file permissions to access any devices
or directories that any remote user will need to write to.
To specify the remote hostname on which the command will be executed, you can
specify either a hostname or an IP address. The hostname might require the full
domain name to reach the remote host.
Only commands that are specified in the.remote_cmds file on the server can be
executed remotely. Any attempt to execute a remote command not defined in this
file will fail.
The -i flag indicates that the command should read standard input from the user
or that standard input will be piped to the remote command from a local
command. If the -i flag is not specified and input is requested from the remote
command, the command locks up indefinitely. The -i flag can be specified,
however, even if no input is requested, but performance might be affected.
The sbclient command causes the sbserver command to be executed on the remote
system. The sbserver command is a daemon that establishes the socket connection
with the local system, sets environment variables, executes the specified command,
reads input, and writes output to the socket.
In addition to the standard environment variables used for login users, the
following environment variables are automatically set on the remote system before
the remote command is executed:
CLIENT_BDIRS
Specifies one or more backup directories, separated by colons, as defined in the
.remote_access file, to which this host and user can read or write.
CLIENT_DEVS
Specifies one or more devices, separated by colons, as defined in the
.remote_access file, to which this host and user can read or write.
Appendix A. Commands
A-59
CLIENT_HOST
Hostname or full domain name (if required) of client host performing the
command.
CLIENT_HOSTPERM
Indicates the hostname, as it appears in the .remote_access file, that granted
permission to the client. This can be the client hostname, the full domain name
of the client, or all.
CLIENT_IDIRS
Specifies one or more installation image directories, separated by colons, as
defined in the .remote_access file, to which this host and user can read or
write.
CLIENT_IPADDR
IP address of the client host performing the command.
CLIENT_USER
User name on the client host performing the command.
CLIENT_USERPERM
Indicates the user name, as it appears in the .remote_access file, that grants
permission to the client. This is either the same as CLIENT_USER or all.
The above environment variables can be accessed by the remote applications
running on the server and can also be queried from the client using a command
such as:
sbclient Remotehost echo \$CLIENT_DEVS
Flags
-i
Indicates the remote command must receive standard input from the local
system. Use this flag only when executing an interactive command or piping
data to the command.
Examples
1. To read the header of a remote backup tape in device /dev/rmt0 on host saturn,
type:
sbclient saturn sbread -H /dev/rmt0.1
2. To copy the file /tmp/imagefile to the remote host neptune, type:
cat /tmp/imagefile | sbclient -i neptune \
"cat > /tmp/imagefile"
In this example, the local file /tmp/imagefile is sent as input to the sbclient
command. The cat command is executed on the remote system, which copies
the data received into a new file /tmp/imagefile. Note that the new file on the
server will be owned by the sbnet user ID.
Files
/usr/lpp/sysback/.remote_access
Contains list of hosts and users that are allowed to
execute commands on the local system.
/usr/lpp/sysback/.remote_cmds
Contains list of commands that can be remotely
executed on the local system.
A-60
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Related Information
The cfgremsvs, cfgremaccess, sbread, sbwrite, and cat commands.
Appendix A. Commands
A-61
sbcomp and sbuncomp
Purpose
Compresses and decompresses data respectively.
Syntax
sbcomp
sbuncomp
Description
The sbcomp command compresses data from standard input and sends the
resulting compressed data to standard output. The sbuncomp command
uncompresses data from standard input and sends the resulting decompressed
data to standard output.
The sbcomp and sbuncomp commands are used as filters and use only memory to
perform the compression or decompression (there is no intermediate file storage
used). Performing software compression and decompression uses a great deal of
CPU resource and can affect the overall performance of the system. If the data is to
be written to a hardware device that performs its own compression, do not
compress the data before writing to the media, as this uses CPU resources
unnecessarily.
If you select to compress data before writing to media when performing any
SysBack backup command with the -p flag, the resulting backup data is piped
through the sbcomp command before sending it to the sbwrite program, which
places the data on the output media. The SysBack restore commands always read
the backup header to determine if the data on the backup was compressed using
sbcomp, and then pipe the data output by sbread to the sbuncomp program
before the data is restored to disk.
Flags
None.
Examples
1. To compress data in the file /data and send the results to sbwrite to write to
the tape drive /dev/rmt0, type:
cat /data | sbcomp | sbwrite -Svp -tl -nrawlv /dev/rmt0
2. To decompress and restore the file /data from the backup performed above,
type:
sbread -Sv /dev/rmt0 | sbuncomp > /data
Related Information
The sbread and sbwrite commands.
A-62
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
sbcron
Purpose
Adds, updates, lists, or removes a regularly scheduled command.
Syntax
sbcron
-n
-u
-r
-l
-m
-h
-d
-M
-w
″minutes″
″hours″
″day″
″month″
″weekday″
-s ″script″
-D ″description″
Description
The sbcron command is used to add, update, list, or remove a regularly scheduled
command. Regularly scheduled commands are scheduled using one or more of the
following flags: -m, -d, -M, or -w.
The following rules apply to this command:
v You must use a whole number for the -m, -h, -d, -M, and -w flags.
v To indicate an inclusive range, use two numbers separated by a dash. For
example, to schedule a command to run Tuesday through Friday, use the -w flag
with a value of ″2-5.″
v To indicate that a script should be run on separate days (not an inclusive range),
separate the numbers with a comma. For example, to schedule a command to
run on the first and last day of January, use the following flags:
-M 1 -d 1,31
Note: Do not use a space after the comma.
v When you add or update a new entry (created using -n), you must use the -m
and -h flags.
v If the -d, -M, or -w flags are omitted, all allowed values will be used in place of
the missing flag. For example, to schedule a command to run at 6:30 AM every
day of every month, use the following syntax:
sbcron -s /vg_backup -m 30 -h 6
To schedule a command to run at 6:30 AM every Monday of every moth, use the
following syntax:
sbcron -s /vg_backup -m 30 -h 6 -w 1
Flags
-n
Creates a new schedule for the script defined with the -s flag.
-u
Updates an existing schedule.
-r
Removes an existing schedule.
-l
Lists all scheduled SysBack scripts.
Appendix A. Commands
A-63
-m ″minutes″
Specifies the minutes for scheduling a script. Valid values are from 1 to 60.
-h ″hours″
Specifies the hour for scheduling a script. Valid values are 0 to 23, where 0 is
equal to 12:00 AM, and 23 is equal to 11:00 PM.
-d ″days″
Specifies the day a script should be scheduled. Valid values are from 1 to 31.
-M ″month″
Specifies the month a script should be scheduled. Valid values are from 1 to 12,
where 1 is equal to January, and 12 is equal to December.
-w ″weekday″
Specifies the day of the week a script should be scheduled. Valid values are
from 0 to 6, where 0 is equal to Sunday and 6 is equal to Saturday.
-s ″script″
The script to be scheduled.
-D ″description″
Enables you to include a custom description. This description is included in
the list created using the -l flag.
Examples
1. To schedule the vg_backup script:
sbcron -m 0 -h 23 -w 1-5 -s /scripts/vg_backup
2. To schedule the system_backup script to run at 12:00 AM on the first, fifteenth,
and thirtieth of each month:
sbcron -m 0 -h 0 -d 1,15,30 -s /home/scripts/system_backup
3. To schedule the fs_backup script to run at 6:30 AM every Monday, Wednesday,
and Friday:
sbcron -m 30 -h 6 -w 1,3,5 -s /fs_backup
4. To schedule the system_backup script to run every day at 12:00 AM:
sbcron -m 0 -h 0 -s /system_backup
A-64
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
sbdevice
Purpose
Checks tape drive availability and controls tape movement.
Syntax
sbdevice
device
-b
-B
-e
-f ″num″
-n
-q
-r
-s
-v
-w
Description
The sbdevice command is used to check the availability of the specified tape drive
or drives and to perform other tape commands. Either tape drive names or a
virtual device name can be specified for the device parameter.
For tape devices, you only need to specify the base device name (for example,
rmt0). The tape will not be moved unless rewinding, forwarding, or backspacing,
regardless of the tape device names or virtual device name specified.
Flags
-b
Causes a tape drive or virtual device to back up a file mark. If the drive is at
the beginning of a backup image, it will be positioned at the start of the
previous backup image. If the drive is currently within a backup image, it will
be returned to the start of the current image.
-B
Reads and returns the current block size of the tape drive. A value of 512 is
returned for all other device types.
-e Rewinds and ejects tapes from the specified drives. If a sequential autoloader is
used, the next sequential tape is inserted automatically.
-f “num”
Forwards the tape past the specified “num” number of file marks. This is used
to position to the next backup image on the tape.
-n
Indicates that the program is to exit with an error code if any errors should
occur. The default action is to display a message on most errors and wait for
the tape to reset.
-q
Indicates that no errors should be displayed. The command exits with a
non-zero return code if an error occurs.
-r
Rewinds the specified device or devices.
Appendix A. Commands
A-65
-s
Checks if the tape drive is currently at the beginning of the tape. Return code
of 0 indicates the tape is not rewound; 1 indicates the tape is rewound.
-v Indicates that messages (such as “Rewinding /dev/rmt0 ...”) should be displayed
as operations are performed on each tape drive.
-w
Indicates that the device or devices should be checked for write permission.
For tapes, the write-protect switch is checked and an error is returned if it is
set. For files, an error is returned if the user does not have write permission to
the file or directory.
Examples
1. To read the current block size of tape drive /dev/rmt0, type:
BS=vsbdevice -B rmt0v
2. To rewind and eject all tapes from drives included in the vdev0 virtual device,
showing messages as the drives are rewound, type:
sbdevice -ve vdev0
3. To only check for write permission to the file or directory
/usr/lpp/sysback/images/venus.rootvg.01201244:
sbdevice -w /usr/lpp/sysback/images/local/VG.venus.01201244
4. To rewind device /dev/rmt0 and /dev/rmt1, indicating that the command should
not wait if an error occurs, type:
sbdevice -nr rmt0 rmt1
A-66
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
sbejecttape
Purpose
Used to eject either a local or remote tape drive or SysBack virtual device.
Syntax
sbejecttape -f ″device″
-h ″hostname″
Description
Use this command to eject a tape from a local or remote tape drive or from a
SysBack virtual device.
Flags
-f ″device″
Specifies the output device name. The device can be a tape drive or a virtual
device name and can be located on the local system or a server (if the -h flag is
also used).
-h ″hostname″
Specifies the remote tape server host to which the device is attached. This
option is available only for remote backups after both the tape server and the
local machine have been properly configured.
Appendix A. Commands
A-67
sbeot
Purpose
Forwards to the end of the data on a SysBack-created tape.
Syntax
sbeot
-f ″device″
-h ″hostname″
-q
Description
Use this command to forward to the end of the last backup on a tape created by
one of the following SysBack commands: sysback, mkvgback, mkjfsback,
mkdirback, mklvback. You can then append another SysBack backup to the end of
the data on the tape.
The following restrictions apply:
v All the images on the tape must have been created by SysBack Version 4 or
higher.
v The tape on which you want to append a backup must be loaded in the tape
drive. For example, if the previous backup spanned three tapes, you must load
the third tape into the tape drive before running the sbeot command.
Flags
-f ″device″
Specifies the output device name. The device can be a tape drive or a virtual
device name and can be located on the local system or a server (if the -h flag is
also used).
-h ″hostname″
Specifies the remote tape server host to which the device is attached. This
option is available only for remote backups after both the tape server and the
local machine have been properly configured.
-q
Suppresses the output. The command displays error messages if they occur.
Examples
1. To position the tape to end of a local system and suppress progress of
positioning:
sbeot -qf /dev/rmt0
2. To position the tape to end on a remote server:
sbeot -h hercules -f /dev/rmt1
A-68
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
sbfwd
Purpose
Forward one or more tape devices to a specific image.
Syntax
sbfwd
-a
-c ″termdev″
-v
-p
-s ″num″
device
Description
The sbfwd command is used to forward a specified number of images to a tape or
tapes written with the sbwrite program. sbfwd is a low-level command called by
the SysBack restore programs when multiple backups are appended to the same
media.
If a virtual device was used to perform the backup, you can omit the -P and -a
options when specifying a virtual device name to forward. If a virtual device name
was not used, and the backup being read was written with a parallel write-policy,
specify the -P flag and include all devices in the list. If the backup was created
with a sequential write-policy, you can list multiple device names, and the
forwarding continues on the next sequential device when the first one reaches end
of volume.
This command should be used only by experienced users who have created
backups using the sbwrite command outside of the normal SysBack backup
process.
Note: You must specify no-rewind tape device names (such as /dev/rmt0.1) with
this command. Otherwise, the devices are automatically rewound when the
command is complete.
Flags
-a Indicates the device is an auto-loader. You are not prompted to change
volumes, but the forwarding continues automatically when the next volume is
inserted.
-c “termdev”
If the backup contains multiple tape volumes, the volume prompt is sent to the
“termdev” device (such as /dev/tty0).
-v Indicates summary header information should be displayed as each image is
read. This gives the header information of the image that is being skipped.
-P
Indicates the backup was performed with a parallel write-policy, so all
specified devices should be forwarded synchronously. If not specified, a
sequential write-policy is assumed, and only the first device in the list is
forwarded unless the end of media is encountered.
-s “num”
Number of images (backups) to skip
Appendix A. Commands
A-69
Examples
1. To forward devices /dev/rmt0 and /dev/rmt1 in parallel to the third image
(skipping 2 images), type:
sbfwd -P -s2 rmt0 rmt1
2. To forward virtual device vdev0 to the next image, type:
sbfwd -s1 vdev0
3. To forward autoloader device /dev/rmt2 to the fifth image (skipping four
images), showing the header on each backup image as it is passed, type:
sbfwd -av -s4 rmt2
Related Information
The sbread and sbwrite commands
A-70
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
sbread
Purpose
Reads data from all SysBack backup types.
Syntax
sbread
device
-a
-c ″device″
-e
-v
-x
-H
-L
-P
-T
Description
The sbread command is used to read a backup that was created using the sbwrite
command. All SysBack backups are written to the media using sbwrite, and all
SysBack restore programs usesbread to retrieve the data from the media. In
addition, SysBack uses sbwrite to store the table of contents (LVM information),
file list, and other information identifying the type and contents of the backup.
This information can then be retrieved using the sbread command.
The device parameter can be one or more tape drives, a virtual device, or a dash (-)
indicating that the input should be received from standard input (stdin).
The sbread command reads data only from backup devices or directories to which
the local host and user has access, as defined by the cfglocaccess or cfgremaccess
commands. When sbread is executed locally, the access permissions are checked in
the .local_access file. When executed by the sbclient command on a remote host,
the remote host and user access permissions are checked in the .remote_access file.
By default, sbread assumes the data was backed up with sbwrite using a
sequential write policy. If you used a virtual device name to perform the backup,
you can use the same virtual device name to read the backup. Otherwise, if you
are reading from a backup created with a parallel write policy, you must specify
the -P flag to the sbread command. Also, when reading from a backup made using
a parallel write-policy, you must also specify the same number of devices to read
the data as was used to create the backup.
SysBack backups, with the exception of the raw logical volume backups, are
formatted using either the AIX backup or tar command. The resulting data is sent
to the sbwrite program, which stores the data on the media, including the backup
header, table of contents (LVM and filesystem information) and file list, if specified.
To read the same backup, it is necessary to use the sbread program to read the
data from the media and send the results to the AIX restore or tar program to be
unformatted and placed in the AIX filesystems.
When you use the -H, -T, or -L options with a no-rewind tape device name, the
data is read from the beginning of the backup media and the tape is repositioned
at the beginning of the same backup image. By omitting the -H, -L and -T flags, it
Appendix A. Commands
A-71
is assumed you are to read the actual data from the backup image. You can
compress data before writing to the media. If compressed data is written to the
media, the data must be uncompressed before it is restored to disk. The SysBack
backup programs compress data using sbcomp and also indicate to sbwrite that
the backup is compressed by using the -p flag. Before reading a backup with
sbread, you can query whether the data was compressed by reading the backup
header (sbread -H). If Packed: Y is shown, you should uncompress the data as it is
output by sbread using the sbuncomp program.
If the data you want to read is not contained in the image the tape is currently
positioned to, you must use the sbfwd command to forward the tape to the correct
image before reading.
Flags
-a Indicates the device is an auto-loader. This flag suppresses the volume prompt,
requesting that the user change tapes. Instead, a message indicates the tape is
unloading and waiting for next volume. The reading continues automatically
when the autoloader has inserted the next volume.
-c “termdev”
If the backup contains multiple tape volumes, the volume prompt is sent to the
“termdev” device (such as /dev/tty0).
-e Indicates all devices (if tape) should be rewound upon completion. The default
action is to not rewind.
-v Indicates whether status messages should be displayed at the beginning and
end of the backup.
-x Displays the progress indicator when reading the data.
-H
Indicates the header information should be read from the backup. If using a
no-rewind tape device, the tape will be repositioned to the beginning of the
same image when completed.
-L
Reads the file list from the backup, if it exists. If using a no-rewind tape
device, the tape is repositioned to the beginning of the same image when
completed.
-P
Indicates the backup was performed with a parallel write-policy, so all
specified devices should be read synchronously. If not specified, a sequential
write-policy is assumed, and only the first device in the list is read unless the
end of media is encountered.
-T
Reads the table of contents (LVM and filesystem information) from the backup,
if it exists. If using a no-rewind tape device, the tape is repositioned to the
beginning of the same image when completed.
Examples
1. To read the contents of a compressed sequential backup created using the
sbwrite command to virtual device vdev1, saving the results in a file called
/tmp/file, type:
sbread vdev1 | sbuncomp > /tmp/file
A-72
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
2. To read the header (label) on device /dev/rmt1 and to not rewind after reading,
type:
sbread -H rmt0
3. To read the table of contents from the media, type:
sbread -T rmt0 > /tmp/toc.out
4. To read the data contained in the parallel backup of /dev/rmt0 and /dev/rmt1,
displaying the progress indicator as the backup is read, and sending the output
to the AIX restore command to restore the data to the AIX filesystem, type:
sbread -Px rmt0 rmt1 | restore -xvqf-
Files
/usr/lpp/sysback/.remote_access
Contains list of hosts and users that are allowed to
execute commands on the local system.
/usr/lpp/sysback/.remote_cmds
Contains list of commands that can be executed
remotely on the local system.
Related Information
The sbcomp, sbuncomp, sbfwd and sbwrite commands.
The cfglocaccess and cfgremaccess commands.
The AIX backup and restore commands.
Appendix A. Commands
A-73
sbscript
Purpose
Creates, updates, or removes a shell script containing the commands to execute.
Syntax
sbscript
-o
-r
-l
-s ″scriptname″
-D ″description″
Command
-c ″command″
Description
Used to create, update, or remove a shell script containing one or more commands
to execute. When creating or updating a script, you can type the command name
or specify it using the -c flag.
Flags
-o Overwrites existing script name, if it exists.
-r
Removes script name, if it exists.
-l
Lists currently defined scripts.
-s ″scriptname″
Name of the script to create, update, or remove.
-D ″description″
Describes the shell script.
Command
Specifies the command or commands contained in the shell script. Separate the
commands using spaces.
-c ″command″
Specifies the command or commands contained in the shell script. Commands
must be inside double quotation marks.
Examples
Use the following syntax to create a shell script called system_backup that will be
a full system backup:
sbscript -s system_backup -D "System Backup -c "sysback -vf /dev/rmt0 sbvg vg01"
The following script is created:
#!bin/ksh
#
# Description: System Backup
#
sysback –vf /dev/rmt0 sbvg vg02
rc=$?
exit $rc
A file in /usr/lpp/sysback/scripts/.toc_script is also updated. The format of the file
is as follows:
A-74
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Script Name Description
The file created from the above example would be as follows:
system_backup System Backup
To remove the script “wed.backup,” type:
sbscript -r -s /user/lpp/sysback/sbscripts/wed.backup
Appendix A. Commands
A-75
sbspboot
Purpose
Initiates a NIM Resource Network Boot of an SP node.
Syntax
sbspboot
-d
-n
-c
-c
″host″
″spot name″
″lppsource name″
″display″
Description
The sbspboot command is used to initiate a NIM Resource Network Boot of an SP
node. Use this command instead of the Classic Network Boot method to ensure
that the proper SP and PSSP specific scripts are executed. For more information
about this command, see Chapter 13, “Network Boot/Installation Configuration”,
on page 13-1.
This command must be initiated from the SP Control workstation.
The specified NIM SPOT and NIM LPPSOURCE entries must be the same as
those specified when the client was added for network boot using the
mksbnetboot command.
Flags
-h “host”
Specifies the hostname of the SP node to be booted.
-n “spot name”
Specifies the NIM SPOT resource to be used for the network boot process.
-l “lppsource name”
Specifies the NIM LPPSOURCE resource to be used for post-installation
processing.
-d “display”
Specifies the hostname where a console window will be displayed during a
prompted installation.
Examples
To initiate a network boot of an SP node called “bcn1e” using the NIM SPOT
“spot_aix433” and the NIM LPPSOURCE “lppsource_aix433” and to open a
console window on the host “lasher,” type:
sbspboot -h bcn1e -n spot_ais433 -l lppsource_aix433 -d lasher:0
Related Information
The mksbnetboot and sbspotcust commands.
A-76
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
sbspotcust
Purpose
Installs SysBack into a NIM SPOT resource.
Syntax
sbspotcust
″devname″
″dirname″
-l ″lppsource name″
-F
-n ″spot name″
-q
-d
Description
When performing a NIM Resource Network Boot, SysBack must be installed into
the NIM SPOT resource in order for the SysBack programs to be available during
a network boot. The sbspotcust command reads the installation filesets from the
specified input device or directory, copies them into the specified NIM
LPPSOURCE directory, and installs them into the NIM SPOT from the NIM
LPPSOURCE directory.
Note: You must run this command before adding a NIM Resource Network Boot
client to ensure that the client’s boot image contains the SysBack programs.
If you install or update the SysBack filesets contained in the NIM SPOT
resource after the network boot client is added, you must update the
network boot image used by the client using the mksbnetboot command.
Flags
-d “devname | dirname”
Specifies the input device name or directory name that contains the SysBack
filesets.
-F Indicates to overwrite the same or newer version of SysBack in the NIM SPOT
resource with the version contained on the installation media.
-l “lppsource name”
Specifies the NIM LPPSOURCE name to which the filesets will be copied.
-n “spot name”
Specifies the NIM SPOT name to which the filesets will be installed.
-q
Checks the specified spot to determine if a valid level of SysBack is installed.
The -n flag is required.
Examples
1. To install SysBack to the NIM SPOT resource called “spot_433” from the
/dev/cd0 device through the NIM LPPSOURCE called “lppsource_433”, type:
sbspotcust -d /dev/cd0 -n spot_433 -l lppsource_433
2. To query the level of the spot in spot_433, type:
Appendix A. Commands
A-77
sbspotcust -q -n spot_433
Return Codes
0
Successful update of the specified spot
1
Spot does not contain the SysBack filesets.
2
Spot does not contain SysBack filesets.
3
Image location is invalid or does not exist.
4
Invalid spot or lpp_source specification.
5
Update of specified spot failed.
Related Information
The mksbnetboot command.
A-78
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
sbwrite
Purpose
Writes raw data to backup media.
Syntax
sbwrite
-S
-P
-C
device
-a
-b
-B
-c
-d
-e
-i
-n
-p
-r
-R
-s
-t
-T
-v
-x
″size″
″backend″
″device″
″description″
″file″
″contents″
″read_access″
″type″
″file″
Description
The sbwrite command reads data from standard input, adds a backup header,
optional file list, and LVM information, and writes to single or multiple devices
either sequentially, in parallel, or by creating multiple copies. The device parameter
can be either a single tape drive, multiple tape drives, a virtual device, or one or
more disk image file names.
This command should be used only by the experienced user who wants to create
backups without using the SMIT menus, and who is knowledgeable in the use of
the sbfwd and sbread commands needed to restore data from backups made using
this command.
The sbwrite command will write data only to backup devices or directories to
which the local host and user have access, as defined by the or cfgremaccess
commands. When sbwrite is executed locally, the access permissions are checked
in the .local_access file. When executed by the sbclient command on a remote
host, the remote host and user access permissions are checked in the
.remote_access file.
Any command that produces a data stream to standard output can pipe that data
stream to the sbwrite command. The sbwrite program then stores the data on the
media, including the backup header, LVM information, and file list, if specified.
The data can later be read from the backup media by the sbread program, which
can then pipe the data stream to another command.
If the -i file option is specified, the data is read from the specified file, which can
also be a logical volume name (preceded with the /dev prefix), instead of from
standard input. If a dash (-) is specified for the output device name, the output is
sent to standard output (stdout) instead of to a device or file.
Appendix A. Commands
A-79
Unless you are using a virtual device name for the device parameter, specify either
the -S flag (indicating sequential backups), the -P flag (indicating parallel backups)
or -C flag (indicating multi-copy backups). Each of the backup types are described
in Chapter 10, “Virtual Devices”, on page 10-1.
The other options are used by the SysBack backup commands mklvback,
mkjfsback, mkdirback, mkvgback and sysback to define the contents and format
of the data contained in the backup but do not alter the way the data is read using
the sbread command. For example, the -t “type”, -n “description” and -d
“description” flags can be used to supply user and backup-specific information
that will be stored in the header label on the backup.
The type parameter can be one of the following:
s
System backup
p Power system backup
v Volume group backup
L Multi-logical volume backup
F Multi-filesystem backup
l
Single logical volume backup
f
Single filesystem backup
d file/directory backup
o other/unknown. This is the default if none is specified.
The “backend” parameter, specified with the -B flag, can be any single character
that is used by SysBack to keep track of the format of the data in the data stream.
For filesystem data, SysBack uses either “B” for the AIX backup command or “T”
for the AIX tar command.
Data can be compressed using the sbcomp command before sending to the sbwrite
command. If so, you should specify the -p flag to sbwrite, indicating that the data
on the media is compressed, so you will know later to uncompress the data during
a restore.
When sbwrite is executed remotely using the sbclient command, the hostname
and user name of the actual user and host performing the command remotely are
written to the backup header. When executed locally, the local user and hostname,
if any, are included in the header.
To include a filelist on the backup, supply the -L flag, followed by the name of the
file containing the file list. This same option can be used to store any type of data.
To read this data from the backup and display the results to stdout, use the -L flag
of the sbread command. Likewise, to store additional information in the table of
contents portion of the backup, use the -T flag with the name of the file containing
the table of contents data. The table of contents can be read from the backup using
the -T flag of the sbread command.
Flags
-a Indicates the device is an auto-loader. User is not prompted to change
volumes, but the reading continues automatically when the next volume is
inserted.
A-80
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
-b “size”
Specifies the buffer size in K-bytes (1024-bytes). Changing this value either
increases or decreases the amount of data that is written to the output device
at one time. Some devices with small buffers might require you to reduce this
value, while other devices with very large buffers can benefit from increased
performance by increasing this value. If you are unsure, use the default value
(64 Kbytes).
-B “backend”
Specifies any single character representing the command used to create the
data stream.
-c “termdev”
If the backup contains multiple tape volumes, the volume prompt is sent to the
“termdev” device (such as /dev/tty0).
-C
Indicates the backup should be performed with a multi-copy write policy. The
same data is written to all specified devices. If used, multiple device names
must also be provided.
-d “description”
A custom description to be included in the backup header. Up to 60 characters
can be used. This information will be placed in the backup header. If the
description includes spaces, the entire description must be enclosed in double
quotation marks (″).
-e Indicates all devices (if tape) should be rewound upon completion. The default
is to not rewind at end.
-i “file”
Indicates the name of a file or logical volume (with /dev prefix) that contains
the data to be written using sbwrite. If not specified, sbwrite reads the data
from standard input.
-l “level”
Specifies a single character representing the incremental backup level. SysBack
backups use a digit between 0 and 9.
-L “file”
Indicates the file list, included in the file specified by the “file” parameter,
should be included on the backup.
-n “contents”
If supplied, “contents” describes the contents of the backup (such as the
volume group name or directory name). Up to 128 characters can be included.
This information is placed in the backup header. If contents includes spaces,
the entire contents must be enclosed in double quotation marks (″).
-p
Indicates the data being sent to sbwrite is packed (compressed). This is
informational only and does not alter the input data. Specifying the -p flag
with sbwrite causes the “Packed” field in the backup header to show “Y”
when reading the header with the “sbread -H” command.
-P
Indicates the backup should be performed with a parallel write-policy, so all
specified devices should be written to synchronously. If specified, multiple
device names must be supplied.
Appendix A. Commands
A-81
-r user=[u|a]
Specifies user read permission when writing to a disk image file. “u” indicates
only this user can read the file. “a” indicates any user can read the file.
-r host=[h|a]
Specifies host read permission when writing to a disk image file. “h” indicates
only this host can read the file. “a” indicates any host can read the file.
-R
Indicates the file list and table of contents file, if any, are to be removed after
the command is complete.
-S
Indicates the backup should be performed using a sequential write-policy. In
this case, the backup is written to the first device. When full, the next device in
the list is used for the second volume. Only when all specified devices are full
are you prompted to change the media in all drives. This option should be
used by default when only one device name is specified.
-t “type”
Indicates the type of backup being performed. Valid types are d
(file/directory), l (raw logical volume), v (volume group), 0 through 9
(incremental filesystem level) or o (other).
-T “file”
Indicates the table of contents, included in the file specified by the “file”
parameter, should be included on the backup.
-v Indicates summary header information should be displayed before writing the
data. This information is sent to standard error.
-x Indicates the progress indicator should be displayed as the files are being
written. If -x is used, the command sending the data to sbwrite should not
update the screen.
Examples
1. To create a backup of all files using the AIX backup command, compressing the
data with sbcomp and formatting the output to 2 parallel auto-loading tape
drives, rmt0 and rmt1, type:
find / -print | backup -ivqf- | sbcomp | sbwrite -aP \
-td rmt0 rmt1
2. To write the file /home/myfile sequentially to the virtual device vdev1,
including a backup description and displaying the progress indicator while
writing, type:
cat /home/myfile | sbwrite -x -td -n"Transaction log" \
vdev1
3. To write the file list contained in the file /tmp/filelist to a backup of the /home
directory, writing multiple copies to drives /dev/rmt1, /dev/rmt2, and /dev/rmt3,
type:
find /home -print | backup -ivqf- | sbwrite -Cv -td \
-n"/home" -L/tmp/filelist rmt1 rmt2 rmt3
Files
/usr/lpp/sysback/.remote_access
Contains list of hosts and users that are allowed to
execute commands on the local system.
A-82
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
/usr/lpp/sysback/.remote_cmds
Contains list of commands that can be remotely
executed on the local system.
Related Information
The sbclient, sbcomp, sbuncomp, sbfwd and sbread commands.
The AIX backup command.
Appendix A. Commands
A-83
sysback
Purpose
Backs up the entire system or specific volume groups for use in a system
installation.
Syntax
sysback
-v
-x
-f ″device″
-f ″directory″
-f ″file″
-B
-T ″platform″
-k ″kernel″
-b
-c
-d
-e
-p
-g
-h
-k
-i
-I
-r
-M
-n
-N
-O
-P
-T
-U
-X
″size″
″termdev″
″description″
″filename″
″hostname″
″kernel″
″uniqueID″
″read_access″
″nettype″
″platform″
″host″
VGname
-I ″ID″
Description
The sysback command is used to create a backup of the operating system (rootvg
volume group) and optionally other volume groups on the system. This backup
can be used to completely reinstall a system to its original state or to install a new
system from the customized image. It can also be used to recreate or restore select
files, directories, filesystems, logical volumes, or volume groups on an already
active system. If backing up to a tape device, the tape will be bootable and will
include the installation programs needed to install from the customized backup.
The backup can be performed to a tape drive, virtual device, or disk image file.
When specifying a disk image file, you can enter the full pathname of the file to
create or overwrite. When doing so, the filename must begin with the prefix “ SB.”
to indicate the file is a SysBack system backup. You can also specify only the
backup directory and a unique ID. When doing so, a filename will be constructed
for you. Refer to “Backups to Disk Image Files” on page 2-6 for information on the
filename created. If you are creating a file and a file by the same name already
exists, specify the -O option to overwrite the existing file.
The backup can also be performed to a remote device, virtual device, or disk
image file. If Remote Services has been configured on both the local and server
system using the cfgremsvs command, and you have defined a backup device on
A-84
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
the server for use by this system using the cfgremaccess command, then you can
also select a hostname using the -h option. The backup is then written to the
device or filename on the specified server, provided the server has enabled this
system access to the specified device or directory.
The backup can also be performed as a “pull” backup to enable central
management of backups. The machine that “pulls” the backup is called the
“initiator,” while the machine that is backed up is called the “backup system.” The
machine that receives the backup data is called the “destination system.”
The initiator is the machine where the mkdirback command is issued. The backup
system is referenced in the command syntax by the -U “host” flag. The data
destination system is referenced by the -h “host flag”.
When only the -U “host” flag is specified, the -f “device” flag refers to a device on
the initiator system. In this case, the backup destination and the initiator are the
same system. This is known as a “two-way pull backup.”
When the -U “host” and -h “host” flags are specified, the -f “device” flag refers to
a device on the host specified by the -h “host” flag. In this case, the backup
destination and the initiator are different systems. This is known as a “three-way
pull backup.”
Remote Services must be configured, using the cfgremsvs command, on all of the
machines involved in the backup operation, the initiator, the backup system, and
the data destination machines. The backup destination machine must have the
Server Options of Remote Services configured to allow the backup system machine
to send its data to devices on the backup destination machine. You can do this
with the Add or Change Client Host Access to this Server option or the
cfgremaccess command.
The backup system machine must have the Client Options of Remote Services
configured to specify the backup destination machine as a backup host. You can do
this with the Add or Change List of Remote Backup Servers option or the
cfgremserver command.
The backup system must also configure Remote Command Access to allow the
initiator machine to pull the backup.
To include network support in the boot images of the tape, specify the -N nettype
option. This enables a network installation to be performed after booting a system
from this tape. This flag is valid only when a tape device is used. Valid network
types are ethernet, token ring or FDDI. You can also specify a device name (such as
/dev/ent0, /dev/tok0 or /dev/fddi0) for the nettype parameter.
The sysback command calls the mkvginfo program to generate information on
volume groups, logical volumes, filesystems, paging space, and physical volumes.
This information can be used later to recreate the system or any of the included
volume groups to its original state. The mkvginfo command creates a temporary
file that is placed in the table of contents on the backup, or you can create a
customized file using the mkvginfo and editlvminfo commands and include it
instead on the backup by also specifying the -g “filename” option to this
command. Files, or files within directories, that are specified in the exclude list are
not backed up with this command. See the chexcludelist command for information
on creating exclude lists.
Appendix A. Commands
A-85
Cloning Systems: A System backup created on one machine can be installed on
another machine with a different processor, platform type, or other system devices.
However, the machine being backed up must have installed all of the device
support for the destination platform type, processor type, and other required
devices.
If the platform or kernel type differs between the system being backed up and the
system that is installed from the backup, the boot image on the tape must be
created for use on the destination system. This is accomplished by specifying the
-T “platform” and -k “kernel” options.
The following platform types are currently supported with the -T flag:
chrp
Common Hardware Reference Platform
rs6k
″Classic″ RISC System/6000 Uni or Multi-processor (AIX 4.2+)
rs6ksmp
Multiprocessor RISC System/6000 (AIX 4.1 only)
rspc
PCI-based (PC) RISC System/6000
The following kernel types are supported with the -k flag:
up
Uniprocessor (single processor)
mp
Multiprocessor
Power System Backup: If the -P flag is specified, a power system backup is created.
The power backup differs from the regular system backup in that all filesystem
data is backed up as raw logical volume data. This can increase the performance of
the backup and restoration of the data considerably but imposes certain limitations
on the system installation and restoration of data from the backup:
v No filesystem attributes or mount point can be changed during a system
installation from this backup.
v The logical volume name, logical volume size, and stripe size attributes of the
logical volumes cannot be changed during a system installation from this
backup.
v Only entire logical volumes or filesystems can be restored from the backup, and
only if the logical volume and filesystem, if applicable, are currently inactive. It
is not possible to restore individual files or directories from the backup.
v The entire filesytem is backed up as raw logical volume data. This might result
in longer backup and restore time for filesystems that were only partially full.
When you create a system backup, sysback performs the following functions:
1. Executes the user-created Pre-backup Script. Appendix C, “Creating Scripts for
Customizing the System Backup and Install Process”, on page C-1 describes
how to create this script. This script can perform any function required by the
user prior to the backup process.
2. If you are backing up to a tape device, the block size of the tape drive is
temporarily changed to 512-byte blocks, if necessary. This ensures that the
installation programs are able to read from the media. The block size is
changed back to its original value before the volume group files are placed on
the media.
3. Generates an LVM information file, using the mkvginfo command, that will
contain the information for all volume groups, logical volumes, and filesystems.
A-86
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
This file is used by the installation process to regenerate the same volume
group configuration. This step is ignored if the -g “filename” option was
specified.
4. Places boot image and installation programs on the tape in separate images.
This enables the installation process to be accessed by booting from the tape.
This step is ignored if the backup media is not a tape.
5. The data for each filesystem and logical volume is written to the backup media,
each in a separate backup image.
6. Executes the user-created Post-backup Script. This script can perform any
functions required by the user after the completion of the backup, but prior to
the rewinding and verifying of the backup format. Appendix C, “Creating
Scripts for Customizing the System Backup and Install Process”, on page C-1
provides information on creating this script.
Flags
-b “size”
Specifies the buffer size in K-bytes (1024-bytes). Changing this value either
increases or decreases the amount of data that is written to the output device
at one time. Some devices with small buffers might require you to reduce this
value, while other devices with very large buffers can benefit from increased
performance by increasing this value. If you are unsure, use the default value
(64 Kbytes). The value specified must be a multiple of the tape block size.
-B
Indicates that only boot images are to be created on the tape. This flag is
invalid unless a tape device is specified. When this flag is used, all other
options, with the exception of the device name and platform type, are ignored.
-c “termdev”
If the backup contains multiple tape volumes, the volume prompt is sent to the
“termdev” device (such as /dev/tty0).
-d “description”
A custom description to be included in the backup header. Up to 60 characters
can be used. This information will be placed in the backup header. If the
description includes spaces, the entire description must be enclosed in double
quotation marks (″).
-e Forwards to the end of the last SysBack backup on the tape. Used when
stacking backup images on the same tape.
-f “device, file or directory”
Specifies the output device name, filename, or directory for the backup. The
specified output option can be on the local system or server (if -h option
provided). The device name can be a tape drive or virtual device name. If a
directory is specified, you must also specify an ID using the -I flag.
-g “filename”
Indicates the name of a LVM information file previously created and
customized for this backup. Refer to the mkvginfo and editlvminfo
commands.
-h “hostname”
Name of the remote backup server host to receive the backup data (such as the
data destination host). This is the host that provides the backup device
specified by the -f flag.
Appendix A. Commands
A-87
-i
Indicates to dynamically create a backup file IDD when writing backups to a
disk image file. This flag supersedes the -I“ID” flag and is useful when
scheduling a backup to disk image file.
-I “ID”
This option is used only if a directory is specified as the output device. The ID
is included in the filename created to make it unique from other disk image
file backups. If a backup in the same directory with this ID already exists,
specify the -O flag to overwrite the previous backup.
-k “kernel”
When backing up to tape, indicates the kernel type of the machine that will be
booting from the tape. The default is to create a bootable tape for the current
kernel type.
-M
Retains physical partition mapping, by default, when logical volumes are
recreated from this backup. Unless the logical volumes were originally created
using physical partition maps, it is generally recommended that they not be
recreated using the same partitions, as this would preserve fragmentation that
develops over time as logical volumes and filesystems are expanded. You can
specify whether or not partition mapping is used before the logical volumes
are recreated.
-n
Indicates that the backup is to be placed at the current tape location without
rewinding the tape. Only if the tape is positioned at the start of tape are boot
images placed on the tape. If the tape is not positioned at the start of the tape,
the backup is appended to the prior backup and can later be restored from by
entering a backup sequence number with the sysrestore command.
-N “nettype”
Specifies that device support for the indicated network type should be
included on the tape boot images.
-O
Indicates that, if a filename was specified as the output device, and the
filename already exists, that the new backup will overwrite the previous file.
-p
Indicates that the data should be packed before being written to the media.
This typically reduces size of backup between 25% and 40%. For increased
performance, do not use this option when backing up to a device that provides
hardware data compression.
-r user=[u|a]
Specifies user read permission when writing to a disk image file. “u” indicates
only this user can read the file. “a” indicates any user can read the file.
-r host=[h|a]
Specifies host read permission when writing to a disk image file. “h” indicates
only this host can read the file. “a” indicates any host can read the file.
-T “platform”
When writing to tape, specifies the platform type for which the boot image
will be created. The default is to create boot images for the current platform
type.
-U “host”
Name of the host to be backed up (such as the backup system).
-v Lists the files as they are backed up. This flag cannot be used with the -x flag.
A-88
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
-x Displays the progress indicator as the backup is being performed for each
volume group. This flag cannot be used with the -v flag.
-X
Excludes all non-JFS (journaled filesystem) logical volumes, or “raw” logical
volumes, from the backup.
Examples
1. To generate a backup of the root volume group only, to the tape drive
/dev/rmt0, displaying the progress indicator, type:
sysback -nxf rmt0
Because the -n option is specified, the tape needs to be rewound. If the tape is
already at the start of media, the resulting backup will be on a bootable tape. If
not, the backup will be appended to the prior data.
2. To generate a backup of the root volume group and the uservg and datavg
volume groups, using virtual device vdev2, and making the backup bootable on
a multiprocessor RSPC machine, enter:
sysback -f vdev2 -T rspc -k mp uservg datavg
3. To generate a compressed (packed) backup of the rootvg and uservg volume
groups to the disk image file /usr/lpp/sysback/images/all/SB.venus.04281007 on
host mars, listing files as they are backed up, enter:
sysback -pvh mars -f \
/usr/lpp/sysback/images/all/SB.venus.04281007 uservg
4. To initiate a pull backup of the /home filesystem on host lasher, from host
sysback1, and send the data to rmt1 of the host sysback1, type the following from
sysback1:
sysback -U lasher -f /dev/rmt1 /home
5. To initiate a pull backup of the /home filesystem on host lasher, from host
sysback1, and send the datat to rmt1 of the host shappy, type the following
command from sysback1:
sysback -U lasher -h shappy -f /dev/rmt1 /home
Related Information
The mkvginfo, cfgremsvs, cfgremaccess,cfgremrootaccess, and chexcludelist
commands.
Appendix A. Commands
A-89
sysrestore
Purpose
Restores or verifies data on any SysBack backup or lists backup type or contents.
Syntax
sysrestore -f ″device or file″
-c
-D
-F
-h
-s
-L
-n
-T
-v
-y
-z
-D
-F
-L
-R
-V
-f
-l
″termdev″
″destination″
-v
-x
-t
″hostname″
″sequence″
″filelist and path″
data
Description
The sysrestore command is used to either verify or restore data from any type of
backup created with SysBack. This command can also be used to list the backup
type or the contents of the backup.
The command can be performed from a remote device, virtual device, or disk
image file. If Remote Services has been configured on both the local and server
system using the cfgremsvs command, and you have defined a backup device on
the server for use by this system using the cfgremaccess command, then you can
also select a hostname using the -h option. The command can then access devices
or files on the server, provided the server has enabled this system access to the
specified device or directory.
When specifying a local or remote filename, do not include the file extension (.TOC
or .lvname) added by the SysBack backup command. The sysrestore command
automatically reads the specified data from the file or files with the appropriate
extension.
Different types of data can be verified or restored from different types of backups:
A-90
Backup type
Verify/Restore types
System backup
volume groups, filesystems, logical volumes,
directories, files
Power backup
volume groups, filesystems, logical volumes
Filesystem
filesystems, directories, files
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Logical volume
logical volumes
File/Directory
directories, files
When restoring a volume group, all logical volume and filesystem data within the
volume group is restored. When restoring directories, all files within the directories
are restored as well, even if the files are contained in different filesystems.
The data argument can be a list of either volume groups, filesystems, logical
volumes, directories, or files, depending on the type of data to be listed, verified,
or restored (as indicated by the -t flag).
The -t option must be specified to indicate the type of data to verify or restore. For
instance, a volume group backup can contain multiple volume groups, filesystems,
and raw logical volumes. To restore two specific filesystems from the backup, you
must specify the -tF option, and the data argument at the end of the backup must
contain the mount points of the filesystems to restore. To restore only a specific
directory from a backup, regardless of the backup type, you must specify the -tD
option, and the data argument must be the directory or directories to restore.
To verify the backup by reading the data, specify the -V flag. When specified, the
data on the backup is read but no data is restored. You still need to specify the
type and data you want to verify. For example, to verify only a couple of raw
logical volumes on a volume group backup, specify the -V flag, the -tL option, and
the names of the logical volumes to verify as the data argument. To verify the
entire contents of a System backup, you must specify the -V flag, the -tV flag, and
the names of all volume groups to verify.
The -L flag is used to list the contents of the backup, depending on the type
specified with the -t flag. The following table indicates the data list that is created
with the -L flag, based on the type:
type (-t)
Produces list of:
V
Volume groups
L
Logical volumes
F
Filesystems
D
Directories
R
Regular files
f
Directories and regular files
l
Filesystems and logical volumes
The -T flag queries the type of backup and returns a single character as follows:
S
System backup
P
Power system backup
V
Volume group backup
F
Filesystem backup
L
Logical volume backup
D
File/directory Backup
Appendix A. Commands
A-91
The -i “sequence” option is used to specify which backup to verify or restore when
multiple backups are stacked onto the same tape or series of tapes. If this flag is
specified, and the value is greater than 1, the tape or tapes will be automatically
forwarded to the beginning of the desired backup before verifying or restoring
data.
The -F flag must be specified to allow restoration of system directories or files that
are not ordinarily replaceable. Restoring files to these directories can have adverse
affect on the system or even cause a system failure, and should be used with
extreme caution. Those directories are:
v /
v /dev
v /etc/vg
v /etc/objrepos
v /usr
v /usr/lib
v /usr/lib/objrepos
Specifying a new destination: The -D “destination” option indicates that the data
should be restored to the new destination.
When you restore from a file/directory backup, the files or directories will be
restored relative to the specified destination but will include the full path of the
original files. For instance, if a file /home/tony/file1 is restored to destination /tmp,
the resulting file would be /tmp/home/tony/file1.
When you restore from a system, volume group or filesystem backup, the files are
restored to the destination directory, but the filename does not contain the prior
mount point of the filesystem. For instance, if files are restored from the /home
filesystem to the /tmp destination, a file previously called /home/tony/file1 will be
restored as /tmp/tony/file1.
To restore logical volume data to a new location, specify a new logical volume
name. The specified logical volume name cannot be currently active.
Incremental Restores: The -y flag must be specified to enable restoration of
incremental data. To restore the entire contents of an incremental backup, this flag
must be specified. The resulting filesystems will contain the same files as it did at
the time the backup was created. This can, in some cases, require files to be
re-removed from the filesystem that were deleted prior to the incremental backup
level being performed.
If the -y flag is not specified, you can restore only select files and directories from
an incremental backup. Refer to “Understanding Incremental Backups” on page 4-1
for important details on restoring data incrementally.
Notes:
1. This command is not intended for restoring an entire AIX operating system.
The only way to restore the entire system is to reinstall from the system
backup. Refer to Chapter 16, “System Installation and Maintenance”, on page
16-1.
2. When specifying the files to restore, SysBack supports the use of BRE (Basic
Regular Expression) wild cards to restore a wildcard match of files. The files
and the wildcard must be enclosed in double quotation marks (″) to avoid
A-92
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
expansion by the user’s shell. For example, “/home/j*” would indicate to
restore all directories and files that start with /home/j such as /home/john or
/home/jenn.
Flags
-c “termdev”
If the backup contains multiple tape volumes, the volume prompt is sent to the
“termdev” device (such as /dev/tty0). If the -h hostname option is specified,
this flag indicates the device on the remote host.
-D “destination”
Indicates that files should be restored to the directory specified by the
“destination” parameter or the logical volume data should be restored to the
logical volume specified by the “destination” parameter, rather than to their
original locations.
-f “device or file”
Specifies the input device name or filename of the backup. The specified
option can be on the local system or on a backup server (if -h option
provided). The device name can be a tape drive or virtual device name.
-F Enables the restoration of system files or directories otherwise non-replaceable.
-h “hostname”
Hostname of backup server to which the remote backup device is attached.
-H
Creates a detailed usage listing on the screen for this command. When
specified, no other action will be performed.
-i “sequence”
Specifies the backup to list, verify, or restore on a tape or series of tapes
containing multiple (stacked) backups. The default is “1”, or the first backup
on the media.
-L
This flag causes the command to read and display the items on the backup of
the specified type (indicated by the -t flag).
-n
Indicates the tape is already positioned at the beginning of the backup to read.
When this flag is supplied, the -i sequence option is ignored.
-t “type”
Indicates the type of data to be listed, verified, or restored.
-T
Reads the backup header and displays a single character representing the
backup type. No other action is performed.
-v Displays the file names as they are either verified or restored. This flag has no
affect when verifying or restoring logical volumes. Do not specify this flag if
the -x flag is used.
-V
Indicates that the backup contents should be verified, rather than restored.
Only items of the specified type (indicated by the -t flag) are verified.
-x Displays the progress indicator, including the approximate size and time to
read the entire contents of the media and the amount completed. Do not
specify this flag if the -v flag is used.
Appendix A. Commands
A-93
-y Indicates that an incremental restore can be performed. This flag is required if
the backup is an incremental backup and restoring other than specific files or
directories.
-z “filelist and path”
Specifies the path and name of a file containing a list of files to restore.
Return Codes
The sysrestore command returns a value, which can be queried with the $? shell
variable, indicating the success or failure of the process as follows:
0
The process was completely successful.
1
The process terminated due to an unknown error.
2
The process terminated because there was an error writing the backup data
to disk.
3
There were read errors when reading the backup data from the backup
media. The failure resulted in loss of some data and possibly termination of
the restore, depending on the number of read errors and how SysBack is
configured to handle them.
4
There were sync errors when reading the backup data from the backup
media. The failure resulted in loss of some data and possibly termination of
the restore, depending on the number of sync errors and how SysBack is
configured to handle them.
5
The process completed successfully, however there were non-fatal errors
when writing one or more files to disk.
Examples
1. To restore the /tmp directory and its contents from the disk image backup
/usr/lpp/sysback/bf/VG.localhost.06230300, displaying the progress indicator
instead of the file list as the files are read, type:
sysrestore -x -tD \
-f/usr/lpp/sysback/bf/VG.localhost.06230300 /tmp
2. To restore the files /data/file1 and /data/file2 from a backup on virtual device
vdev1, type:
sysrestore -tR -fvdev1 /data/file1 /data/file2
3. To list all files and directories included in the backup on device /dev/rmt0
attached to host neptune, enter:
sysrestore -L -tf -h neptune -f rmt0
4. To show the names of all logical volumes and filesystems included on the
backup on tape device /dev/rmt1, type:
sysrestore -L -tl -frmt1
1. To verify the uservg volume group on the backup on device /dev/rmt0, showing
the files as they are read, type:
sysrestore -Vv -tV -f rmt0 uservg
2. To determine the backup type of the second backup on the tape, type:
sysrestore -T -f rmt0 -i2
Related Information
The sysback, mkdirback, mklvback, mkjfsback and mkvgback commands.
A-94
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Appendix B. Booting a System for SysBack Installation or
Maintenance
This appendix provides the specific step-by-step instructions for booting a system
to the SysBack Installation and Maintenance menu, used to install a system from a
SysBack system backup.
The instructions for booting the system to the Installation and Maintenance menus
differ depending on whether the system is booted from a bootable tape or from a
network boot server, and also differ depending on the type of the machine to be
booted. Instructions are provided for each of the following machine types:
v Common Hardware Reference Platform (uni or multiprocessor)
v Microchannel-based RISC System/6000 (uniprocessor)
v Microchannel-based RISC System/6000 SMP (multiprocessor)
v PCI-based (RSPC) RISC System/6000 (uni or multiprocessor)
v Scalable POWERparallel (SP) Node (uniprocessor)
v Scalable POWERparallel (SP) Node (multiprocessor)
If an error occurs during a network or tape boot, a value might appear in the
system LED panel. On many PCI-based systems, the LED panel is optional but
makes troubleshooting a network boot much easier. Refer to Appendix E, “LED
Codes”, on page E-1 for a detailed explanation of the LED codes used in SysBack.
For LED values that are not listed in Appendix E, refer to the AIX documentation.
There are three steps to boot a system:
1. Initiate the boot process.
2. Initiate the network boot (for systems booted from the network).
3. Complete the boot.
Initiating the boot process
The following sections describe how to begin the boot process for the different
types of systems.
Common Hardware Reference Platform (uni or
multi-processor)
The CHRP systems are actually PCI-based and the boot process is identical to other
PCI-based systems. To boot these systems, refer to the instructions in “PCI-based
(RSPC) RISC System/6000 (uni or multi-processor)” on page B-3.
Microchannel-based RISC System/6000 (uniprocessor)
You can boot a microchannel-based RISC System/6000 system in one of two ways:
from a SysBack system backup tape or from a network boot server.
Booting from a SysBack system backup tape
To boot the system using a system backup tape:
1. Insert the system backup tape in the drive.
2. Turn the system key to the SERVICE position.
© Copyright IBM Corp. 2002
B-1
3. Turn on the system or press the yellow button twice.
4. Continue with “Completing the Boot Process” on page B-4
.
Booting from a network boot server
Use these instructions to obtain the BOOTP screen:
1. Turn the system key to the SECURE position.
2. Turn on the power to the system. After a short sequence of numbers, the LED
displays 200.
3. Turn the key to the SERVICE position and press the yellow reset button once.
4. If the LED value displays 260, 261, or 262, and the BOOTP Main menu is
displayed on the screen, continue with “Initiating the Network Boot” on
page B-4.
If the above instructions did not work, you need to boot from an IPLROM
Emulation Diskette:
1. Insert the IPLROM Emulation Diskette.
2. Turn the system key to the SERVICE position.
3. Turn on the power to the system or press the yellow reset button twice.
4. Continue with “Initiating the Network Boot” on page B-4.
Microchannel-based RISC System/6000 SMP (multiprocessor)
You can boot a microchannel-based RISC System/6000 system in one of two ways:
from a SysBack system backup tape or from a network boot server.
Booting from a SysBack system backup tape
1. Insert the system backup tape.
2. Power off the system.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
Turn the system key to the SERVICE position.
Press Enter until the “>” prompt appears.
Type sbb and press Enter to view the STAND-BY.
Select option 1, Set Flags.
Select option 2, Autoservice IPL. Make sure this value is set to “enable”. If so,
press “n”. If it is currently set to “disable”, press “y” to change it.
Select option 6, Fast IPL. Make sure this value is set to “enable.” If it is, press
“n”. If it is currently set to “disable,” select “y” to change it.
Type x to return to the “>” prompt.
Type x again to clear the screen.
Power on the system.
Continue with “Completing the Boot Process” on page B-4.
Booting from a network boot server
1.
2.
3.
4.
5.
6.
B-2
Power off the system.
Turn the system key to the SERVICE position.
Press Enter until the “>” prompt appears.
Type sbb and press Enter to view the STAND-BY.
Select option 1, Set Flags.
Select option 2, Autoservice IPL. Make sure this value is set to “enable.” If it
is, press “n”. If it is currently set to “disable,” press “y” to change it.
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
7. Select option 6, Fast IPL. Make sure this value is set to “enable”. If it is, press
“n”. If it is currently set to “disable,” select “y” to change it.
8. Type x to return to the “>” prompt.
9. Type x again to clear the screen.
10. Power on the system.
11. From the MAINTENANCE menu, select option 6, SYSTEM BOOT.
12. Select option 1, BOOT FROM NETWORK.
13. Press Enter at the language prompt for English. The BOOTP MAIN menu is
displayed.
14. Continue with “Initiating the Network Boot” on page B-4.
PCI-based (RSPC) RISC System/6000 (uni or multi-processor)
For PCI-based RISC System/6000 (RSPC) machines, booting from tape or network
is accomplished through the System Management Services (SMS). System
Management Services is included in the firmware of most models and is loaded
from diskette on other older models. System Management Services can be executed
using a graphical, icon-based, interface, or an ASCII, text-based, emulation. The
steps provided below are for the ASCII (text-based) interface.
Booting from a SysBack system backup tape
1. If the RSPC model does not include SMS in the firmware, insert the System
Management Services Diskette.
2. Power on the system.
3. When the first screen appears (a picture is displayed and icons begin showing
at the bottom of the screen), press F4.
4. From the System Management Services menu, select Select Boot Devices and
press Enter.
5. Select Boot Other Device.
6. Select the tape drive to boot from.
7. Continue with “Completing the Boot Process” on page B-4.
Booting from a network boot server
1. If the RSPC model does not include SMS in the firmware, insert the System
Management Services Diskette.
2. Power on the system.
3. When the first screen appears (a picture is displayed and icons begin showing
at the bottom of the screen), press F4.
4. From the System Management Services menu, select Utilities and press Enter.
5. Select Remote Initial Program Load Setup.
6. Select IP parameters.
7. Fill out all fields, including the server IP address, gateway address and subnet
mask.
8. Press Esc twice to return to the top menu.
9.
10.
11.
12.
Select Select Boot Devices.
Select Boot Other Device.
Select the network adapter to boot from.
Continue with “Completing the Boot Process” on page B-4.
Appendix B. Booting a System for SysBack Installation or Maintenance
B-3
Initiating the Network Boot
From the BOOTP Main menu, use these steps to initiate a network boot:
1. Select option 1, Select BOOT1, Select BOOT (Startup) Device.
2. Do not select the Default Boot (Startup) Device option, but rather the
appropriate network adapter and connector that applies to your node. Make
sure you select the appropriate interface option (BNC, DIX or built-in ethernet,
4 or 16 MB token-ring data rate).
3. Change the Client address, Server address and Gateway address as needed by
selecting the number corresponding to each entry field. Then, enter the IP
address for each. The IP address entries requires leading zeros in each portion
of the IP address (i.e. “005.001.001.024”). The Gateway address may be set to the
Server’s IP address if both systems are in the same segments.
4. When all entries are complete, type “99” to return to the MAIN MENU.
5. Select the option to Exit Main Menu and Start System (BOOT). The option
number will vary.
6. On SP Systems, put the key in NORMAL mode using the Node Front Panel.
Press Enter once more to continue.
7. Continue with the “Completing the Boot Process”.
Completing the Boot Process
If booting from a network boot server and the client is configured for no-prompt
install, the installation proceeds without any user prompts. Refer to the section Set
Network Install Client Defaults in Chapter 9, “Remote Services”, on page 9-1 for
details on the configuration of a no-prompt install.
Unless this is a no-prompt install, the following prompt will appear on the screen:
****** Please define the System Console ******
Type 1 at this terminal and press <Enter>
if you want this display to be the System Console.
The above example is for an ASCII terminal. Note that the prompt Type 1 is Press
F1 for graphical displays. If the console is an ASCII terminal, press 1 and Enter. If
the console is a graphic display, press F1 and Enter.
After a short time, the following SysBack Installation and Maintenance screen is
displayed:
B-4
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
IBM Tivoli Storage Manager for System Backup and Recovery
Installation & Maintenance
MAIN MENU
Change Installation Device
Change Volume Group & Logical Volume Information
Install the System with Current Settings
Utilities Menu
Reboot the System
+------------------------------------------------------------------------------+
| Installation Device: Network [ent0]
|
|
|
+------------------------------------------------------------------------------+
| Use ARROW or TAB keys to move and ENTER to Select.
|
+------------------------------------------------------------------------------+
The detailed instructions for using the options on this menu are provided in
Chapter 16, “System Installation and Maintenance”, on page 16-1 and Chapter 18,
“Changing the Volume Group, Logical Volume and Filesystem Attributes”, on
page 18-1.
Troubleshooting a Network Boot
To boot from the network, a server must have first been configured to provide a
network boot image to this client. Those steps are outlined in Chapter 13,
“Network Boot/Installation Configuration”, on page 13-1 under Create a Network
Boot Image and Configure a Network Boot Client.
The system hardware must either be BOOTP enabled, or you must have created an
IPLROM Emulation Diskette to perform a network boot. All PCI and SMP systems
are BOOTP enabled, but some older Microchannel-based (non-SMP) systems are
not. If you are unable to obtain the BOOTP MAIN MENU using the prior steps,
your system might not be BOOTP enabled.
Determining the Network Adapter Hardware Address
Configuring the network boot client sometimes requires the network adapter
hardware address of the client. The steps to obtain the network adapter hardware
address of the client differ depending on the type of platform:
1. Microchannel-based/SP systems: The network adapter address is determined
by performing the first steps in a network boot. Follow the steps below for
initiating a network boot on this platform. When selecting the network adapter
to boot from, the hardware address of the adapter is displayed.
2. PCI-based (RSPC) systems: The network adapter address may be displayed
using System Management Services (SMS). SMS is included in the firmware of
some rspc models, and is loaded from diskette on other models. To determine
the hardware address, follow these steps:
a. If the RSPC model does not include SMS in the firmware, insert the System
Management Services Diskette.
b. Power on the system.
Appendix B. Booting a System for SysBack Installation or Maintenance
B-5
c. When the first screen appears (a picture is displayed and icons begin
showing at the bottom of the screen), press F4.
d. From the System Management Services menu, select Utilities and press
Enter.
e. Select the Tools icon.
f. Select Remote Initial Program Load Setup.
g. Select Adapters.
h. Select the network adapter the machine will boot from.
i. Record the Adapter Hardware Address.
B-6
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Appendix C. Creating Scripts for Customizing the System
Backup and Install Process
To add additional flexibility and automation to the backup and installation process
for requirements specific to a particular customer environment, SysBack can call
user-created shell scripts, if they exist, before and after the system backup and after
completion of the installation process when installing from a system backup.
These shell scripts can perform any function required and are executed with root
user authority. The following are examples of functions you might want to
perform:
Pre-backup functions:
v Remove all temporary files from the system to prevent unnecessary
backup time and space.
v Send a message to all users that a backup is being performed.
v Automatically log users off the system.
v Shutdown the network daemons or disable tty ports to prevent user
logins.
Post-backup functions:
v Restart network daemons and enable tty ports to allow users to log on.
v Send a message to a user or users that the backup is complete.
Post-install functions:
v Enable a shell for the install user to perform maintenance tasks.
v Remove unwanted configuration data from the system (i.e. user IDs and
passwords). This is needed when cloning systems from a single backup
image.
Script Names
During installation of System Backup & Recovery for AIX, the
/usr/lpp/sysback/scripts directory is created. This directory has read and write
permissions only for the root user. Do not give any other user permission to add
or change the scripts in this directory, because all scripts are executed with root
user authority.
You can create four scripts, which can contain any function, in this directory. The
scripts must be owned by root and must have execute permissions.
The SysBack system backup looks for the following scripts and executes them
before installation if they exist:
/usr/lpp/sysback/scripts/fsback.pre
This script is executed when the System Backup option is selected from
the SMIT menus or when the sysback command is executed. The script is
executed before any other function is performed. Any environment
variables exported before the backup is run are available to this script. In
addition, the following variables will be defined:
DEVICE
© Copyright IBM Corp. 2002
Name of tape devices, virtual device, or backup image file.
C-1
HOSTNAME
Name of remote backup server, if any.
/usr/lpp/sysback/scripts/fsback.post
This script is executed at the end of the SMIT System Backup option or at
the end of the sysback command. It is executed after all data is backed up
for the root and other volume groups (if selected), but before the backup
format is verified. Because the tape is left positioned at the end of the
backup, this script might contain commands to append to the tape
additional information that is not part of the system backup. Any
environment variables exported before the backup is run are available to
this script. In addition, the following variables are defined:
DEVICE
Name of tape devices, virtual device, or backup image file.
HOSTNAME
Name of remote backup server, if any.
The SysBack installation process looks for the following scripts and executes them
at the end of installation if they exist:
/usr/lpp/sysback/scripts/install.pre
This script must exist on the system where the system backup tape was
created or on the network boot server. It is restored and executed before
the installation menus appear. The following variable will be defined:
BOOTDEV
Name of the boot device (i.e. “rmt0” for tape or “ent0” for
ethernet adapter).
/usr/lpp/sysback/scripts/install.postroot
This script must exist on the system where the system backup was created.
After all files are restored onto the newly installed system, this script exists
on the new system. It is executed immediately after the root volume group
data is restored, before any additional processing, such as recreating the
boot logical volume or restoring the non-rootvg volume groups. Refer to
“Post-Installation Scripts” on page C-3 for additional information.
/usr/lpp/sysback/scripts/install.post
This script must exist on the system where the system backup was created.
After all files are restored onto the newly installed system, this script exists
exist on the new system. It is executed as the last step in the installation
process, just before the system reboot. Refer to “Post-Installation Scripts”
on page C-3 for additional information.
Note: The scripts must have execute permissions or they will be ignored. To set
the execute permissions, type the following after creating the scripts:
chmod u+x filename
(where “filename” is the name of the script file)
If any of the scripts return a non-zero error code or contain syntax errors, the
backup or installation process alerts you and asks you to press Enter before
continuing. The remainder of the process will then continue as usual. Within the
scripts, checks should be made to ensure they are performing as planned, and the
scripts should exit with a non-zero return code if they fail to perform their desired
function. A “1” return code is sent by executing the exit 1 command from within
the script.
C-2
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Post-Installation Scripts
The post-installation scripts can perform any commands available to the restored
system. They can also perform remote commands, but only to the server from
which the system was installed. There are several environment variables available
to the post-installation scripts. They are as follows:
CLIENTIP
IP address of the restored system, if installed from a network
install server. Note that this must be used in place of the
HOSTNAME variable since the hostname is not defined at this
time.
DEVICE
Local tape drive names such as ″/dev/rmt0″ or remote tape drive
name or filename.
GATEIP
IP address of the gateway machine, if any.
NETDEV
Device name of the network adapter (ent*, tok*, or fddi*), if booted
or installed from a network installation server.
POLICY
Indicates the sequential or parallel backup policy (S/P) if multiple
local tape devices are used.
REL
AIX Release level of the restored system
REMOTE
Set to “1” if the installation was performed from a network install
server. Otherwise, this variable is not set. If set, the /etc/hosts file
on the client contains the “instserver” alias, which can be used
with either the sbclient or other TCP commands to access the
install server.
SERVERIP
IP address of the installation server, if any. You cannot use the
server’s hostname because it is not defined at this time.
SUBNET
Subnet mask, if any.
VER
AIX Version of the restored system.
VGINFO
Filename of the temporary volume group information file. See the
“mkvginfo” on page A-49 for the contents.
You can use these environment variables to access the local or remote installation
devices.
Sample Scripts
There are a few sample scripts in the /usr/lpp/sysback/scripts directory. These can
be modified as desired and enabled by changing the name of the sample script to
the appropriate file name described above.
Note: The sample scripts provided are available to assist users in creating their
own custom backup and installation scripts and are not supported by IBM.
The following sample scripts are contained in this directory:
fsback.pre_killusers
This sample script, when enabled, informs all users on the system that a
system backup will start in 5 minutes. It resends the message each minute
until the 5 minutes expire and then ends all user processes, except for the
root user’s. The system backup continues only after the user processes
have been terminated.
Appendix C. Creating Scripts for Customizing the System Backup and Install Process
C-3
To enable this script, change the name to fsback.pre.
install.postroot_hwmigrate
This sample script allows you to install a system backup created on a
different system onto a new machine, even though the new machine has
different device support software requirements. The new system must be
booted from media that contains the required base device support, but the
source system from where the backup was created does need to have the
software support pre-installed.
To boot the new system from the media with the required bass device
support, place an AIX Version 4 Installation CD in the drive of the new
system during the installation process. After the rootvg volume group data
is restored, the additional required device support is automatically added
to the new system from the CD-ROM.
To enable this script, change the name to install.postroot.
install.post_rmnet
This script automatically removes any network configuration from the
system at the end of the installation process. This is necessary when
installing multiple systems from the same system backup to prevent
multiple machines from existing on the network with the same network
configuration.
This script also prevents SysBack from being removed from the system at
the end of the installation process.
If this script is not enabled, and a system is installed from a system backup
created from a different machine, you are prompted to select whether or
not to remove the network configuration and the SysBack product.
To enable this script, change the name to install.post.
C-4
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Appendix D. Device/System-Specific Information
This chapter provides various notes and tips for using SysBack with various types
of devices, networks, or systems. The devices described in this chapter are not the
full list of devices supported by SysBack. The information provided here is specific
to devices or systems commonly used with SysBack, in which the performance or
usability might be impacted by the configuration of SysBack, the device, or the
system itself.
IBM 7208 8mm Tape Drives
All of the 7208 8mm tape drive models use a physical hardware blocking factor of
1024 bytes. This does not prevent the tape drive block size from being set to 512
bytes, but doing so causes the remaining 512 bytes of each physical block to go
unused. Therefore, half of the tape capacity, and up to half of the backup
performance, is lost if the tape block size is set to 512 bytes.
For best performance, the tape block size should be set to 1024 bytes, or any
multiple thereof.
IBM 3490, 3570, 3590, 7205 and 7337 Tape Drives
These are very high performance tape drives and contain a very large data buffer.
To achieve their highest throughput, send data to the tape drive’s buffer in very
large blocks. To do so, set the Buffer Size to 256 Kbytes when performing any
SysBack backup. Change the block size of the tape drive to either 0 or 262144 (256
Kbytes) using the Change Tape Drive Characteristics option. Doing so helps
ensure the drive continues streaming data without having to stop and wait for
additional data from the system. You can set the buffer size to a higher value as
long as the value is a multiple of the tape drive’s block size.
Note that increased performance of the tape drive does not necessarily result in
faster backups if the data being backed up cannot be read at the same rate the tape
drive is writing. In many cases, reading fragmented data from a filesystem on even
the fastest disk drive can have trouble keeping up with the write performance of
these tape drives. To achieve the best performance, read performance of the disk
drives can be increased by keeping data contiguous on the disk and striping data
across multiple disks to spread the I/O workload evenly across multiple devices.
Also, make sure the tape drive Autoload feature is set to “no.” You can change this
feature using the Change Tape Drive Characteristics option. If this option is set to
“yes,” the loader changes tapes automatically, making the entire set of tapes
appear like a single tape to SysBack. Instead, turn this option off so that SysBack
handles the tape changes, placing a new volume label on each tape.
Other Tape Drives
Any tape drive that is recognized by AIX should be usable by SysBack, although
no support can be guaranteed. Most tape drives have at least a 64 Kbyte buffer,
which is the default buffer size that SysBack uses. If you are using a tape drive
that does not have a large buffer, usually resulting in a “write: I/O error” or “illegal
parameter” messages, you might need to reduce the buffer size value SysBack uses
each time a backup is performed. Try using “32” (Kbytes) as the next value, then
© Copyright IBM Corp. 2002
D-1
“16,”, and so on. If the error continues, either the tape drive or tape media is
defective, or it is not supported by AIX or SysBack.
IBM 7331 8mm Tape Library
If the 7331 8mm Tape Library is installed with 2 tape drives, SysBack backups
usually achieve twice the performance of a single 8mm tape drive when
configured as described in this section.
To configure SysBack and the 7331 library for the best performance, create a
SysBack parallel virtual device containing both tape drives in the unit. Then set the
7331 to “Split Sequential” mode. This allows SysBack to “stripe” the data across
both tape drives, achieving twice the performance, while allowing the 7331 to
change the tapes in each drive as the backup reaches the end of volume.
When in Split-Sequential mode, the 7331 assigns the first, or top, 10 tapes in the
magazine to the top tape drive and the bottom 10 tapes in the magazine to the
bottom tape drive. Note that the SCSI connection to the 7331 library itself is not
used by SysBack, only the SCSI connection to the tape drives themselves.
IBM 7332 4mm Tape Library
The 7332 is a low-cost tape drive providing fast 4mm tape write performance,
while allowing unattended operation of a tape autoloader. When configured for
use with SysBack, either 4 internal or 12 external tapes can be written to
sequentially without user intervention.
To do so, create a sequential virtual device containing the single 4mm tape drive
and indicate that an autoloader is used. Any backups performed to this tape drive
automatically span across as many tape cartridges are in the magazine without
user intervention.
Other Tape Libraries or Autoloaders
Most tape libraries, or autoloaders, that have a “sequential mode” can be used
with SysBack. “Sequential Mode” indicates that the tape loader itself automatically
inserts a new cartridge each time a cartridge is ejected from the drive. The tape
cartridges are usually inserted in sequential order, depending on their location in
the unit, but some loaders allow the user to select the order in which the tapes will
be used.
To use SysBack with an autoloader, create a sequential virtual device containing
the tape drive (not the loader itself), and indicate in the virtual device
configuration that an autoloader is used. Then, set the autoloader to sequential
mode.
By not using any device-specific loader controls, SysBack can work with virtually
any tape loader set in sequential mode. SysBack does not control the loader in any
way. Instead, SysBack rewinds and ejects the current tape from the drive any time
a read or write operation reaches end of tape and waits for the loader to insert the
next volume. Upon doing so, the backup continues automatically without user
intervention.
Because libraries or loaders can differ, SysBack cannot claim to support every
loader, as the sequential operation is the function of the loader and not of SysBack.
D-2
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
It is therefore up to the user to perform the steps to properly configure the loader
and verify the proper operation of the sequential mode as described above.
IBM 7133 Serial Storage Architecture (SSA) disk subsystem
The serial storage architecture provides a high-speed interconnection technology
for SSA disk drives. The drives and cables are hot-plugable, which enables a disk
drive to be replaced while the unit is operational. Also, the location within the
loop is transparent to AIX, enabling a disk not currently in use to change locations
without affecting the operation of the system.
Because AIX does not keep track of the physical location of the disk drive, it is not
possible for SysBack to record and later reference a disk by its physical location.
Instead, SysBack keeps track of the physical volume ID, which is unique to each
individual disk drive. When you recreate a volume group on the same physical
disks as they were previously installed, all disks are recognized and automatically
added to the volume group. However, when you clone systems or replace a drive,
the disk or disks appear to be missing. You can the use the Edit volume group
and logical volume attributes option to re-select the drives for each volume group.
IBM 7135 RAIDiant Array
The IBM 7135 is not a supported boot device by AIX. During the SysBack system
installation process, AIX recognizes the disks as “Other SCSI Disk”. AIX cannot,
however, create a volume group on the disks at this time. The 7135 should be
powered off or disconnected when using SysBack to install the system.
When the installation is complete and before rebooting the system in normal mode,
power on the 7135. Each of the 7135 disks is then recognized as a “7135 Disk Array
Device”, assuming the 7135 device support was installed on the system. You can
then recreate and restore volume groups on the 7135 disks using the SysBack
system backup.
When AIX is installed, it “walks the bus,” configuring all disks in the order in
which they appear. Powering off the 7135 during installation of the rootvg ensures
that the internal disk drives, onto which the rootvg volume group is installed, are
named with the first hdisk names (e.g, hdisk0).
IBM 7137 and Other RAID Devices
RAID storage units have a number of unique features that vary depending upon
the actual RAID box and have an impact on the use of SysBack. RAID boxes
provide fault tolerance in case of a disk failure (except for RAID-0, which is simply
a form of disk striping). Most implementations of RAID use RAID-5 or RAID-1,
which provide fault tolerance for a disk failure.
Before a RAID box can be used, it must be configured. This often includes
choosing the RAID level, number of LUNs (known as “hdisks” to AIX), size, and
other characteristics. Creating and configuring the LUNs is done in various ways:
through software on the RS/6000 such as for the IBM 7135, through a keypad on
the RAID box such as for the IBM 7137, or through other means such as a terminal
attached to the RAID box, switch settings, and so on. Only after the LUNS are
configured in the system do the “hdisks” become available for use in AIX volume
groups.
Appendix D. Device/System-Specific Information
D-3
Some RAID devices, including the 7137, are recognized by AIX without the need to
configure any software on the system; however, the disk descriptions show as
“Other SCSI disks”. In most cases, as with the 7137, the disks can be used by AIX
with no further software configuration. Because no special software configuration
is required, they can be recognized by the SysBack installation process and can be
included in any volume group, including the rootvg volume group.
If the RAID device requires special device drivers or software customization before
they can be used, they are either unrecognized or unusable during the SysBack
installation process. If this is the case, power off the unit during the installation
process to ensure that the remaining system disks receive the first available
“hdisk” names. After the rootvg volume group is installed on other supported
disks, you can power on the RAID unit before rebooting the system in normal
mode. At that time, assuming the RAID software support is installed, the RAID
disks can then be configured into the system. Additional volume groups can be
recreated and restored from the SysBack system backup onto the RAID disks.
D-4
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Appendix E. LED Codes
This chapter contains a list of the system LED codes that might appear when you
boot or install from a SysBack System Backup. Many of the LED codes are for
information only and usually occur when there is no system console available;
others indicate an error in the boot or installation process.
Many of the LED codes appear only during a no-prompt installation. This is because
there may not be a system console attached and the LED is used to provide
information to the user on the progress of the installation, or because an error
occurred that requires user intervention.
The LED codes below are those that are provided by SysBack. Any other codes
that appear are not part of the SysBack product but normally represent a system
software or hardware problem. If the LED code is not listed here, refer to your AIX
software diagnostics documentation for details.
000
Failed to identify the system type during a tape boot
The bootinfo -B command failed due to device support missing from the
boot image that is required to support the client system’s hardware. The
problem usually is caused by missing SCSI device support. You need to
recreate the boot tape from a system containing all device support required
for the system to be installed.
c04
An unrecoverable error occurred during a no-prompt installation
A message has been displayed on the system console. Press Enter on the
console to display the detailed error message.
c06
Failed to identify the system type during a network boot
The bootinfo -B command failed due to device support missing from the
boot image that is required to support the client system’s hardware. The
boot server is likely missing device support for either the token-ring,
ethernet or FDDI adapter. You need to install the appropriate device
support on the server system, reboot the server, and remake the network
boot image for the client.
c21
Unable to configure logical network device during a network boot
The ifconfig command failed. Report the problem to your network service
organization.
c31
Prompting user to select a system console
c32
User selected a graphical display as a console device
c33
User selected an ASCII display as a console device
c40
Installation data is being extracted from the install media during a
no-prompt install
c43
Error restoring the AIX install programs during a tape boot
Either the tape media is defective, the tape drive is defective, or the boot
tape is incomplete. Create another boot tape and try booting from tape
again.
c45
© Copyright IBM Corp. 2002
Failed to configure a system console
E-1
The cfgcon command failed because of missing device support for either
the graphic adapter or standard I/O slot for ASCII monitors. You must
install the appropriate device support on the system from which the
backup tape was made, or install the appropriate device support on the
network boot server, reboot the server, and remake the network boot image
for the client.
c46
Normal installation processing during a no-prompt install
c47
Unable to create a PVID on a disk
This normally indicates a hard disk failure. If you are installing on the disk
for the first time, check and reformat, if necessary, the disk in diagnostics
and retry the installation.
c48
Prompting user for input during a no-prompt install
An error has occurred during a no-prompt installation. This LED is used to
alert the user to the message displayed on the system console. The
installation process has defaulted to prompt mode. Press Enter on the
console to re-display the specific error message, then use the normal
installation menus on the screen to correct the problem and begin the
installation.
c50
The rootvg volume group is being created during a no-prompt install
c52
Changing from boot (RAM) to restored rootvg filesystems during a
no-prompt install
200
The system key is in secure mode when booting
If booting from network, turn the key to service mode and press the reset
key. If booting from tape, power off, turn the key to service, and power on.
260, 261 or 262
The BOOTP menus have been invoked
The system is to be booted from the network and the BOOTP menus have
been invoked. If they are not displayed, press Enter on the console.
510
Configuring base system devices during a tape boot
511 then blank screen
Console configured but messages cannot be displayed
The cfgcon successfully configured a system console device; however, the
console messages are not displayed because of missing or incorrect device
support for the graphic adapter. You must install the appropriate device
support on the system from which the backup tape was made, or install
the appropriate device support on the network boot server, reboot the
server, and remake the network boot image for the client.
512
Configuring additional system devices during a tape boot
513
Restoring installation files from tape during a tape boot
600
Starting a network boot
602
Configuring base system devices during a network boot
603
Configuration of base system devices failed during a network boot
There is likely device support missing from the boot image which is
required to support the client system’s bus, I/O or disk hardware. You
need to install the appropriate device support on the server system, reboot
the server, and remake the network boot image for the client.
E-2
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
604
Configuring the boot network
605
Invalid network boot device provided
The boot server is likely missing device support for either the token-ring,
ethernet, or FDDI adapter. You need to install the appropriate device
support on the server system, reboot the server, and remake the network
boot image for the client.
606
Configuring the boot network.
608
Copying network boot information file from boot server
609
Failed to copy network boot information from boot server
The tftp command failed to tftp the file /tftp/Client Hostname.sbinfo from
the boot server. One of the following usually apply:
1. The file does not have read permissions for all users.
2. The file does not exist.
3. An /etc/tftpaccess.ctl file exists on the boot server but does not provide
access to the /usr/lpp/sysback/netinst/boot directory.
4. On a PCI-based system, you cannot use the leading zero’s for any of
the IP address in the SMS Menu. Even though the first tftp command
works, the second one run will fail.
Rebuild the network boot image on the server and add the network boot
client, making sure all values are correct.
610
NFS mount of /usr filesystem from the server is hanging
The /etc/exports file does not have permissions for the client to mount
/usr. Type exportfs on the server and look for the line starting with
“/usr”. Make sure the client hostname is included on the line and that
/usr is exported with root permission.
Example of output:
/usr -ro,root=node1,node2
In the above example, only node1 and node2 are exported. Add the
network boot client again and verify that all the values are correct for the
server’s IP address, the gateway, and the subnet mask for the client.
611
NFS mount of /usr filesystem from the server failed
This often means that /usr filesystem was not exported to the client (see
LED 610). There are known problems in AIX Version 4.1 that will affect
SysBack network boots. These problems are fixed in the following file set
levels:
v
v
v
v
v
bos.adt.prof 4.1.4.5
bos.rte.libc 4.1.4.5
bos.net.nfs.client 4.1.4.4
bos.sysmgt.quota 4.1.3.1
bos.net.tcp.client 4.1.4.4
Make sure the above filesets are at this level or later. If updating the
filesets, reboot the server, then remake the network boot image on the
server.
613
Configuration of gateway failed during network configuration
Appendix E. LED Codes
E-3
The route command failed to add the gateway address provided. Report
the problem to your network service organization.
E-4
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Index
Numerics
3490/3590 9–track tape drives D-1
7133 SSA disk subsystem D-3
7135 RAIDiant array D-3
7137 RAIDiant array D-3
7208 8mm tape drive D-1
7331 8mm tape library D-2
7332 4mm tape library D-2
BOOTP 13-1, B-1, E-2
menus B-4
bosboot 16-12
buffer
network data 15-16
TCP 15-16
buffer size 4-8
C
A
access
local user 8-1
remote host 9-2
access control list 15-14
activity logging 11-1
changing log size 11-1
viewing entries 11-1
AIX 4.2 18-15
AIXwindows 2-1
allocation group 18-14
changing size 18-15
allow
host access 9-6
user access 8-3, 9-6
auto varyon, volume group 18-4
autoloader 10-3, 10-4, 16-4
installation devices 16-4
notes D-2
B
backing up the installation image 4-3
backup
backing up the installation image 4-4
files or directories 4-24
filesystem 4-14
logical volume 4-19
post-backup script 4-9
pre-backup scripts 4-9
scheduling 7-1
SMIT Menus 2-2
volume group 4-9
backup command 15-14
backup file ID 4-6
backup sequence number
listing backup 5-1
recreate 6-3
restore 6-8
system installation 16-13
verifying backup 5-2
backups
incremental backups 4-1
performing 4-1
restore 6-6
backups, performing 4-1
bad block relocation 18-12
boot, logical volume type 18-10
booting
by machine type B-1
© Copyright IBM Corp. 2002
cfglocaccess A-10
cfgremaccess A-12
cfgremrootaccess A-2
cfgremserver A-15
cfgremsvs A-17
cfgvirdev A-18
chexcludelist A-20
chimagefile A-4
chinstclient A-6
classic network boot 13-2
client, remote access
adding 9-4
listing 9-7
removing 9-7
cloning
system backup 4-3, 15-4
compressing data 4-7
compression
filesystem 18-14
concurrent, volume group 18-5
configuration options, SMIT menus 2-2
copies, logical volume, changing 18-9
cron 2-9
D
deny
host access 9-6
user access 8-3, 9-6
description
changing disk image file 15-7
destination, restore 6-10
device image file
changing 15-7
remote 9-4
verifying 5-4
device name
backup 4-5, 5-2, 15-4
device selector screen 2-4
disk image files 2-6
local user permission 8-4, 9-6
remote user permission 9-6
dump device 6-2, 16-14, 18-10
E
edit
logical volumes, during recreate 6-5
volume groups, during recreate 6-5
editlvminfo A-22
eject, tape 14-3
error, handling 15-15
ethernet, installation from 16-5
exclude list 4-7, 4-10, 4-15, 12-1
adding entries 12-2
listing 12-3
removing 12-4
SMIT menus 2-3
F
fastpath, SMIT 2-9
FDDI, installation from 16-6
file list, listing 5-2
files or directories
backing up 4-24
filesystem
changing attributes 18-12
changing attributres 18-1
data backup command 15-14
removing 6-1
fragment size 18-14
G
getlvminfo A-23
H
host
changing file owner 15-8
changing read permission 15-7
read permission 4-8
hostname
backup 4-5, 5-1
restore 6-4
I
images directory 4-5
import, volume group 18-4
incremental backups 4-1
Installation and Maintenance
menus 15-4
installation image
backing up 4-3
backup file ID 4-6
buffer size 4-8
compressing data 4-7
device name 4-5
exclude list 4-7
forwarding or rewinding tape 4-7
host read permission 4-8
hostname 4-5
images directory 4-5
kernel type 4-6
network installation support 4-7
non-JFS volumes 4-7
non-rootvg volume groups 4-9
X-1
installation image (continued)
physical partition mapping 4-8
platform type 4-6
power backup 4-5
pull backups 4-9
remote volume prompt 4-8
report output type 4-6
user description 4-8
user read permission 4-8
installing, Sysback 1-1
inter-policy 18-11
intra-policy 18-10
J
jfs 4-7, 4-12, 6-2
jfslog 6-2, 18-9
K
kernel type
bootable tape 15-3
L
large files, enabling support 18-15
LED codes 16-2, E-1
list, backup contents 5-1
logical volume
backing up 4-19
changing attributes 18-1, 18-7
changing name 18-8
create 6-2
removing 6-1
Logical Volume Manager (LVM) 15-9
customized installation diskette 15-11
information file 15-9
lscfginfo A-27
LVM information file 13-8
M
maintenance shell 16-11
microchannel, booting B-1, B-2
migratepv command 18-12
mkdirback A-29
mkjfsback A-33
mklvback A-37
mksbnetboot A-41
mkvgback A-45
mkvginfo A-49
mount point, changing 18-10, 18-13
multiprocessor 4-6, 15-5, 16-2
booting B-1, B-2
MWC, changing 18-12
N
NBPI 18-14, 18-15
network
LED codes E-1
network boot 13-1
adding or changing a client
13-11
classic 13-1, 13-2
X-2
13-3,
network boot (continued)
configuring image 13-3
NIM resource boot 13-1, 13-9
removing a client 13-9
setting client defaults 13-5
SP node boot and installation 13-13
troubleshooting B-5
Network Boot Configuration, SMIT
menus 2-3
network install image, defining directory
access 8-4, 9-6
network installation
adding support to boot tape 4-7, 15-5
network installation server
adapter hardware address,
determining B-5
changing install server 16-4
removing settings 16-10
network settings, changing 16-4
ethernet 16-5
FDDI 16-6
token-ring 16-5
NFS, network boot problems E-3
NIM resource boot 13-1, 13-2, 13-9
adding or changing a client 13-11
removing a client 13-13
SPOT resource 13-10
NIM SPOT resource
querying 13-11
no command 15-16, 15-17
no-prompt install 16-1
error codes E-1
non-JFS volumes 4-7
P
paging 6-2, 18-10
parallel, virtual device, installing
from 16-4
permission
host read 4-8
host read, changing 15-9
local user 8-1
remote host 9-2
user read 4-8
user read, changing 15-9
physical partition
changing size 18-5
physical partition mapping 4-8
physical volumes
maximum 18-11
selecting for volume groups 18-2,
18-5
platform type
bootable tape 15-3
system reboot 16-15
post-installation configuration 16-13
power backup 18-9
change installation settings 18-2
changing installation settings 16-7
LVM information file 15-9
recreate options 6-6
system backup option 4-5
priority, logical volume 18-10
progress indicator
example 2-8
pull backups
hostname 4-9
Q
quorum
checking 18-5
R
RAID devices D-3
read error, handling 15-15
readsbheader A-53
recovery install 16-1, 16-12
relocatable, logical volume policy 18-12
remakevg A-55
remakevg command 15-2, 18-1
remote backup devices, SMIT
selection 9-4
Remote Services 9-1
configuring 9-4
SMIT menus 2-3
remote volume prompt 4-8, 5-4
reorgvg command 18-12
report output type
during backup 4-6, 5-4
during restore 6-10
example 2-9
restore
data type 6-8, 6-10
from backup 6-7
rfc1323 15-16, 15-17
rootvg 4-2, 4-4, 4-9, 4-10, 15-12, 16-1,
16-9, 16-11, 16-12, 16-15, 18-4, C-2, C-4,
D-3, D-4
RS/6000 Scalable POWERParallel (SP)
assumptions 17-2
initiating installation 17-3
installation 17-1
NIM 17-2
prompted and unprompted
installations 17-4
troubleshooting 17-4
S
sbalog A-58
sbclient A-59
sbcomp A-62
sbcron A-63
sbdevice A-65
sbejecttape A-67
sbeot A-68
sbfwd A-69
sbread A-71
sbscript A-74
sbspboot A-76
sbspotcust A-77
sbuncomp A-62
sbwrite A-79
scheduled backups 7-1
scripts
creating C-1
post-backup 4-9
post-install 15-13, C-1
post-rootvg 15-12
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
scripts (continued)
pre-backup 4-9, C-1
pre-install 15-12
server
adding 9-7
listing 9-9
removing 9-9
SMIT 2-1, 9-4
accessing 2-1
command to execute 2-9
device selector screen 2-4
fastpath 2-9
file list limitation 6-8
help screens 2-9
output screen 2-7
tape drives 14-1
smit.log file 2-7
SP 16-3, B-1
sparse files 15-15
SPOT resource 13-10
stripe size, logical volume 18-11
sync error, handling 15-15
sysback A-84
removing 16-9
Sysback
activity logging 11-1
SysBack
documentation, accessing 1-3
installing 1-1
installing into a SPOT resource 13-10
performing backups 4-1
prerequisite tasks 1-1
sysback.log file 2-8
sysrestore A-90
system
backing up 4-3
system backup 4-3, 6-2
customizing C-1
excluding files 12-1
installation from 16-1
network install image directory 8-4,
9-6
system data, restoring 6-10
system dump
copying 16-14
system installation 16-1
changing device 16-3
error handling 16-2
parallel virtual device 16-4
system, backup 16-1
system, procedure 16-8
System Management Services,
diskette B-3
T
TCP, tuning performance 15-17
token-ring
installation from 16-5
trusted computing base (TCB) 15-14
type, logical volume 18-10
U
uniprocessor 4-6, 15-5, 16-2
user
adding access 8-2
changing file owner 15-9
changing read permission 15-9
listing access 8-4
local access 8-1
read permission 4-8
removing access 8-4
user description 4-8
utilities 15-1
utilities, SMIT menus 2-4
V
verify
backup content 5-2
data type 5-4
virtual devices 10-1
adding 10-3
multi-copy 10-2
parallel 10-2
removing 10-5
sequential 10-2
SMIT menus 2-3
volume numbering 10-2
volume group
backing up 4-9
changing attributes 18-1, 18-4
changing name 18-9
create 6-2
moving logical volumes
between 18-9
removing 6-1
selecting physical volumes 18-2
volume groups
non-rootvg 4-9
volume numbering, virtual devices 10-2
W
wildcard
in exclude list 12-1
listing file to restore 6-8
write policy, virtual devices 10-4
write-verify 18-12
tape
forwarding 4-7
rewinding 4-7
tape drives 14-1
changing 14-2
configuring 14-2
listing 14-1
SMIT menus 2-3
unconfiguring 14-2
tar command 15-14
tcp_sendspace 15-16
Index
X-3
X-4
IBM Tivoli Storage Manager for System Backup and Recovery: Installation and User’s Guide
Printed in U.S.A.
GC32-9076-00
Download PDF
Similar pages