ADSM - General Disaster Considerations


About this document
Related documentation
Recovery Procedure

About this document

This document will provide a starting point for ADSM and Tivoli Storage Manager (TSM) administrators who wish to develop their own disaster recovery plan. As an administrator, you can use this document as a checklist to follow when you create your own plan for disaster recovery testing. Once you develop your own plan, this document will not be needed again until you decide to revise the plan. We also highly recommend testing your plan before implementation.

NOTE: This document is an informal list of steps, not a formal, supported procedure. No guarantees are made with respect to the information in this document.

This document applies to ADSM versions 3.1 through 4.1 and to any operating system that is supported by an ADSM or TSM server, including AIX versions 4.x. This information focuses on the AIX operating system, however, you can adapt the examples to any supported platform. Visit TivoliŽ Storage Manager Products Technical Support for up-to-date information.

Related documentation

The AIX and product documentation library is available:

The Tivoli TSM documentation can be obtained through the following link.


It is very important that you are familiar with your own processes for disaster recovery and not rely entirely on someone else to provide you with assistance during a disaster. Your designed and tested recovery procedure should be documented very clearly and in great detail. Additionally, your procedure should be updated and retested each time your operations center changes in any signifigant way.

The restoration of clients is faster if selectives are run periodically or if collocation is enabled, which will require the use of many more tapes.


This document assumes that you have a working knowledge of ADSM/TSM and system administration. The list below details items that you should have available at the disaster recovery site. If you need assistance with creating or acquiring the items below, it is recommended that you contact a technical specialist or consultant for assistance. You can also take advantage of the TSM Administration training from Tivoli Education.

NOTE: The server binary directory in this list is specified as /usr/lpp/adsmserv/bin. Please replace this directory with the appropriate directory for your installation. Many times this is the /usr/tivoli/tsm/server/bin directory.

Recovery procedure template

The following steps describe the recovery procedure in an overview format and is not all inclusive. Always test your procedure before a disaster and fill in any missing portions on your own final plan.

    For an SP/2 System

  1. Restore the Control workstation from tape. See SP/2 documentation and support for assistance.

  2. If the ADSM/TSM server is a node, restore the adsm server node from mksysb image on CWS via NIM, or from local tape, depending on your installation requirements.

  3. If restore from system backup is not possible, re-install the operating system and ADSM.

  4. Rebuild the filesystems required by ADSM. This varies by platform. See documentation for your operating system for assistance.

  5. Rebuild the ADSM log volumes and copies, enter:
         dsmfmt -m -log /log/file/location <size in MB>
  6. Rebuild the ADSM db volumes and copies, enter:
         dsmfmt -m -db /db/file/location <size in MB>
  7. Format the primary log and db volumes with:
         DSMSERV FORMAT <# of logs> logvolnames <# of dbs>
    NOTE: Do not specify the copies yet.

  8. Ensure that all hardware devices are available to your operating system.

  9. Set up the ADSM device drivers by following the directions in the TSM Administrator's Guide for your operating system type.

  10. Make any device changes needed to the device config file (devconfig.dsm in your server bin directory).

  11. Restore the adsm database with the command:
  12. Bring up ADSM by entering the command:
    NOTE: Watch console messages for errors.

  13. Once the server is up, enter an administration client command line and enter:

  14. For each copy pool:
  15. For each primary storage pool:

  16. Clients can now begin restoration by recreating disk based storage pools:
         DEFINE VOLUME /path/to/new/file
  17. Remake the Log Copies.
         DEFINE LOGC /primary/log/filename /secondary/log/filename
  18. Remake the database copies.
         DEFINE DBC /primary/db/filename /secondary/db/filename
  19. Restore the primary storage pools.
         RESTORE STG stgname

If you have problems with any of the information in this document, call your support center.

[ Doc Ref: 96713107223758     Publish Date: Jan. 22, 2001     4FAX Ref: 8798 ]