Summary
If a Filer volume is deleted without all of its backup instances being expired from the DPX Catalog, Catalog Condense jobs will result in error return code 4.
Symptoms
Catalog Condense jobs will encounter error return code 4 if a Filer volume is deleted when all backup instances have not yet expired from the catalog. The DPX Catalog will attempt to delete the snapshots that were backed up to this volume and, since the Filer volume cannot be reached, the designated snapshots cannot be deleted.
The following warning messages will appear in the condense job log:
<ww.xx.yy.zz> sssvh Tue Jun 10 08:05:33 2008 SNBSVH_541W Skipping deletion of snapshots on volume(/vol/testvol) of node(<FilerName>). Reason(Volume does not exist)
<ww.xx.yy.zz> sssched Tue Jun 10 08:26:18 2008 SNBSCH5705J Set job status: job BE_CONDENSE (sched_id 3102432) failed (rc = 4)
Resolution
Before deleting a volume from the Filer, we recommend that you un-map all mapped snapshot instances originating from this Filer volume and expire backup jobs that have this volume selected as the destination. Once you have un-mapped the appropriate snapshots and cancelled the affected backup jobs, the Filer volume can be deleted and Catalog Condense jobs can be successfully run. If you miss any of these steps, the DPX Catalog will need to establish communication with Filer on this deleted volume.
If a Filer volume is deleted without the appropriate backups expiring from the Catalog, you will continue to encounter the error message described above. In this situation, the system administrator will need to manually delete cataloged instances of backups that had the deleted Filer volume as destination.
Care must be taken to ensure that no extra backup instances are expired from catalog. Contact Catalogic Technical Support to discuss the procedure for systematically purging entries pertaining to the deleted Filer volume from Catalog.