NSF databases restore to alternate server and alternate partition

To reduce demands on the Domino production server, restore data to an alternate server.

A restore operation involves two steps. First, the backup copies of the databases are retrieved for the IBM Storage Protect server. Second, the recorded transactions in the log files are applied to the databases. If the transaction log files required to recover the databases are archived, they can be retrieved from the IBM Storage Protect server. These steps can have an impact on performance in the processor, and in disk input and output.

The transaction log directory must be on a dedicated physical disk drive for optimal performance. When a dedicated physical disk drive is used, the Domino server can write transactions sequentially to the log. Using this disk is faster than writing transactions to random nonsequential parts of a disk. If the restore operation is run on a Domino production server, the restore and application of the transactions interferes with sequential writing of transactions to the log. Performance of the Domino server is affected, and the time that is required to run the restore operation increases. The application of the transaction logs competes for processor cycles with the Domino server.

Restore operations must be run on an alternate server or on an alternate partition for these reasons.

An alternate server restore is the preferred method since the restore operation has no impact on the performance of the production Domino server. However, the production Domino server and the server on the alternate partition can use separate disk drives for their transaction log directory. If separate disk drives are used, the production Domino server access to the transaction log is not affected by the restore operation on the alternate partition.