IBM Tivoli Storage Manager, Version 7.1

Backing up primary storage pools

To protect client data, back up your primary storage pools to copy storage pools. Copy active client backup data in primary storage pools to active-data pools for fast client restores.

About this task

Tip: Backing up storage pools requires an additional 200 bytes of space in the database for each file copy. As more files are added to the copy storage pools and active-data pools, reevaluate your database size requirements.

Each of the commands in the following examples uses four parallel processes (MAXPROCESS=4) to perform an incremental backup of the primary storage pool to the copy storage pool or a copy to the active-data pool. Set the MAXPROCESS parameter in the BACKUP STGPOOL command to the number of mount points or drives that can be dedicated to this operation.

Procedure

To back up data in a primary storage pool to a copy storage pool, use the BACKUP STGPOOL command. For example, to back up a primary storage pool named ARCHIVEPOOL to a copy storage pool named DISASTER-RECOVERY, issue the following command:
```
backup stgpool archivepool disaster-recovery maxprocess=4
```
The only files backed up to the DISASTER-RECOVERY pool are files for which a copy does not exist in the copy storage pool. The data format of the copy storage pool and the primary storage pool can be NATIVE, NONBLOCK, or the NDMP formats NETAPPDUMP, CELERRADUMP, or NDMPDUMP. The server copies data from the primary storage pool only to a copy storage pool that has the same format.
Restriction: A BACKUP STGPOOL command does not back up a shred storage pool to a copy storage pool unless you explicitly permit it by setting the SHREDTONOSHRED parameter to YES. If this value is not specified, the server issues an error message and does not allow the backup. If this value is specified, the server does not issue a warning when the BACKUP STGPOOL command for the shred pool is run.
Tip: To further minimize the potential loss of data, you can mark the backup volumes in the copy storage pool as OFFSITE and move them to an offsite location. In this way, the backup volumes are preserved and are not reused or mounted until they are brought on-site. Ensure that you mark the volumes as OFFSITE before you back up the database. To avoid marking volumes as offsite or physically move volumes:
- Specify a device class of SERVER in your database backup.
- Back up a primary storage pool to a copy storage pool or associated with a device class of SERVER.
To copy active data, use the COPY ACTIVEDATA command. For example, to copy active data from a primary storage pool named BACKUPPOOL to an active-data pool named CLIENT-RESTORE, issue the following command:
```
copy activedata backuppool client-restore maxprocess=4
```
The primary storage pool must have a data format of NATIVE or NONBLOCK. Copies from primary storage pools with any of the NDMP formats are not permitted. The only files copied to the CLIENT-RESTORE pool are active backup files for which a copy does not exist in the active-data pool.

Results

Because backups and active-data copies are made incrementally, you can cancel the processes. If you reissue the BACKUP STGPOOL or COPY ACTIVEDATA command, the backup or active-data copy continues from the point at which the process was canceled.

Restrictions:

If a backup is to be made to a copy storage pool and the file exists with the same insertion date, no action is taken. Similarly, if a copy is to be made to an active-data pool and the file exists with the same insertion data, no action is taken.
When a disk storage pool is backed up, cached files (copies of files that remain on disk after being migrated to the next storage pool) are not backed up.
Files in a copy storage pool or an active-data pool do not migrate to another storage pool.
After a file is backed up to a copy storage pool or a copy is made to an active-data pool, the file might be deleted from the primary storage pool. When an incremental backup of the primary storage pool occurs, the file is then deleted from the copy storage pool. Inactive files in active-data pools are deleted during the process of reclamation. If an aggregate being copied to an active-data pool contains some inactive files, the aggregate is reconstructed into a new aggregate without the inactive files.

Feedback