Issuing the ABACKUP command with the EXECUTE parameter

Issue the ABACKUP command with the EXECUTE parameter when you are ready to have your data sets actually backed up. Aggregate backup always performs verification processing, as described in the previous topic. The difference is that after verification processing completes successfully, aggregate backup proceeds to back up your data sets.

Aggregate backup directs all messages to the aggregate backup activity log. The aggregate backup starting and ending messages are also listed in the command activity log and displayed to the user who issued the command. When processing completes, determine if the ABACKUP command completed successfully with a return code of zero. Review the activity log for messages ARC6060I and ARC6061I for the volume serial numbers of the aggregate backup control and data files, ARC6071I for the volume serial number for the instruction/activity log file, and ARC6062I for the volume serial numbers for the tapes containing ACCOMPANY data sets. You will need these volumes later when you perform aggregate recovery.

If the ABACKUP command failed, as indicated by a nonzero return code, review the activity log for error messages. Correct any errors and reissue the ABACKUP command. The command always starts processing from the beginning. Repeat this process until the ABACKUP command completes successfully with a return code of zero. If the command fails, the data on the tape is invalid.

Since aggregate backup allows backing up of data sets that have been migrated to ML2 tapes, it is possible that the tape volumes containing the data sets to be backed up reside on an ML2 tape volume that is in use by another DFSMShsm function. See Tape migration level 2 volume contention for more information. If ABACKUP determines that it cannot allocate a tape volume because it is “in use”, it retries the allocation up to 90 times, at 20 second intervals. This takes approximately 30 minutes.

If the preceding retry scenario has been completed and the ML2 tape is still “in-use,” DFSMShsm issues message number ARC6254A to the operator by way of a write to operator with reply (WTOR). The operator then has the option of canceling the aggregate backup or retrying the allocation retry logic.

Aggregate backup will copy multiple migrated data sets from a specific ML2 tape with a single mount, except when a data set spans to another tape.