StorageTek 312564001 The procedure in each case is summarized below, Loss of entire database

Archive Manager database recovery may be required in the following circumstances:

1.loss of an individual tape volume from with an Archive Manager tape database (eg) due to media failure

2.loss of an entire Archive Manager database (eg) due to loss of data center.

In both the above cases, standard Archive Manager database recovery procedures should be followed. A full description of these procedures may be found in the ‘Recovery Control Processing’ section in the Archive Manager User Manual.

The procedure in each case is summarized below:

1.Loss of individual tape volume(s)

Identify the dataset sequence number(s) of the tape volume(s) which have become unavailable. A tape volume’s dataset sequence number may be obtained from the name of the dataset contained on that volume. Refer to “Archive Manager database backup and recovery” on page 89 for a description of Archive Manager dataset naming standards.

Set each sequence number (or range of sequence numbers) in recovery status, using the standard Archive Manager database administration dialog for the database containing the unavailable volumes. The copy to be used for recovery (‘B’, ‘C’ or ‘D’) is established during this process. Archive Manager will automatically access the appropriate backup copy when retrieving objects from all volumes which have been placed in recovery status.

The primary copy of all volumes which are in recovery status should be re-created at a subsequent time using the Archive Manager database recovery utility OTASP130. This utility will automatically remove a volume from recovery status after its primary copy has been re-created. Volumes may be manually removed from recovery status at any time via the Archive Manager database administration dialog.

2.Loss of entire database

If an entire database has been lost (due to hardware malfunction or loss of data center), the database must be placed in ‘disaster recovery’ status via the standard Archive Manager database administration dialog. The copy to be used for disaster recovery (‘B’, ‘C’ or ‘D’) is specified during this procedure.

When a database is in disaster recovery status, all retrievals by Archive Manager from that database will be performed from the appropriate backup copy. In order to guarantee the continuance of established service levels for DB2 applications, all volumes in that copy of the database should be available for automated processing.

DB2 Manager User Guide	91
StorageTek Proprietary