Folks, I asked a question of TSM development regarding Active Data Pools (ADP) and received this excellent explanation. I think you will find it interesting. The one thing that escaped my attention in the documentation is the preference for restore choosing an ADP if one exists rather than a primary or standard copy pool. Thanks, Kelly J. Lipp VP Manufacturing & CTO STORServer, Inc. 485-B Elkton Drive Colorado Springs, CO 80907 719-266-8777 [EMAIL PROTECTED]
________________________________ Active data pools (ADPs) represent a new type of storage pool, in addition to primary and copy pools. ADPs were designed to meet the following requirements. 1. Improved restore performance. ADPs can be used with FILE storage pools to allow rapid client restores. If available, files will preferentially be restored from a FILE ADP rather than a primary or copy pool. There is also some benefit when performing server operations such as GENERATE BACKUPSET. 2. Reduced resources for maintaining copy pools. In addition to being used for client restores, ADPs can also be used for storage pool or volume restore operations. Since an ADP only contains active data, the number of tapes is smaller than for a copy pool. Many customers want to be able to move tape volumes with only active data offsite for DR purposes and ADPs allow them to do so. By storing only active data, the number of offsite tapes is reduced as is the effort to manage those offsite tapes. 3. Reduced size of disk staging pools. Keeping only active versions of data in an ADP reduces the size of the disk pool as compared to keeping both active and inactive data in a copy pool. 4. Reduced data movement. This requirement stems from customers needing to stage data before performing client restore. Staging data required active and inactive data to be moved to disk before the store began. By keeping only active data, ADPs can eliminate the need for staging data to disk before restore. As an alternate design, TSM development also considered implementation of active-only primary pools. This implementation would not have supported restore of storage pools or volumes from the active-only primary volumes nor would it have facilitated movement of those active-only volumes to an offsite location. There were also significant technical issues involving migration (inactive data could not be removed from the active-only pool until it had been migrated) and with storage pool backup (backup from an active-only primary pool to a copy pool could mean that inactive files were not stored in the copy pool).
