I have a specific job that has now become stuck on Job Status "Active: Queued - Verify" while backing up to Disk a total of 5 seperate times with the same issue, although each has occured unpredictibly, on different days, weeks apart.
The job will hang in Verifty indefinitely until I manually cancel, and then the duplicate-to-tape job runs, but also hangs at the same Byte Count where the Verify job stopped (during the duplicate phase of the tape job, not the verify phase), and I have to manually cancel that as well. Ultimately this means we do not have a complete backup of this job on tape.
Digging deeper into the Backup 'Job Activity' log: the Backup phase includes a full set of B2D media names, and the Verify phase stops listing B2D media at about 3/4 of the Backup media list. (The B2D media names are different for each of the job failures, so it's not an issue with a specific B2Dxxxxxx.bkf file.)
The job errors have occured with at least 2 weeks to a month between them; not consistent. I have also removed and re-presented this specific disk to BE in between the last two job errors, for unrelated reasons.)
Running the Debugger revealed that indeed the B2D file that would come next in the Verify sequence was somehow caught up in the PVLSVR process:
735 PVLSVR 2940 6/2/2016 9:09:04 AM 4448 AdammSession::MountSpecificMedia()
Requested Media = {F39F96B9-A134-45F2-B819-12C99A497E4D}
736 PVLSVR 2940 6/2/2016 9:09:04 AM 4448 AdammSession::MountSpecificMedia() - currently mounted
Media = {F39F96B9-A134-45F2-B819-12C99A497E4D}, "B2D321858"
737 PVLSVR 2940 6/2/2016 9:09:04 AM 4448 PvlSession::MountSpecificMedia() - ERROR = 0xE000810E (E_PVL_MEDIA_NOT_AVAILABLE) - 0000 seconds, 0000 (8364) SQL seconds
738 PVLSVR 2940 6/2/2016 9:09:04 AM 4448 AdammSession::Execute( ADAMM_SESSION_EXECUTE_MOVER_MOUNT_SPECIFIC )
ERROR = 0xE000810E (E_PVL_MEDIA_NOT_AVAILABLE)So the PVLSVR is unable to mount B2D321858 for some reason, perhaps because it thinks it is *already* mounted?
Looking at the details of B2D321858, there is nothing remarkable. No read/write errors, 21 seeks, freshly created for the backup job.
The only vaguely related solution item I found was this article (https://www.veritas.com/support/en_US/article.000038222) ... but that discusses Tape media specifically, and there is apparently no option to Retire disk media in the same way. Regardless, the problem is not related to a specific B2D file, as this job has hung in the same way on different B2D files each time.
And I don't think it's related to the content of the backup job either, because a different file is listed in the Activity Log each time the job has hung.
I'll restart the BE services, then do an Inventory/Catalog on the disk in question, but without any predictibility of the job failing, it is hard to isolate and experiment. And interestingly, no other job of our 50+ has this problem: This is backing up a specific directory on a physical server, but the job settings are very standard and similar if not identical to other physical server jobs.
Perhaps this is a database issue?
Any insights or suggestions would be most appreciated.
Thank you.
