VTL - Write Filemark Table failed during relabel of virtual tape

Asset ID:	1-72-1386061.1
Update Date:	2011-12-21
Keywords:

Solution Type Problem Resolution Sure

Solution 1386061.1 : VTL - Write Filemark Table failed during relabel of virtual tape

Applies to:

Sun StorageTek VTL Plus Storage Appliance - Version: 2.0 - Build 1656 and later [Release: 2.0 and later ]
Information in this document applies to any platform.

Symptoms

Receiving write errors when relabeling vtapes.

In VTL console event log:
Write Filemark Table failed.BC [VA###], VID 1000####, EC -6

In VTL message file:
Nov 26 20:38:52 stkvtl05 sanidmgr[673]: [ID 702911 daemon.notice] IPSTOR||0||E||0x00009d00||WriteFMTable Failed||VA0200||10001520||-6

Cause

Write block list failed due to problem rewriting header after shrink operation:
Nov 26 20:58:30 stkvtl05 ipstorcomm: [ID 702911 daemon.notice] IPSTOR||1322359110||I||0x000036bd||succeeded to shrink tape||Virtual Tape barcode [VA0437] (id:10001510)||40960
Nov 26 20:58:30 stkvtl05 genunix: [ID 459543 kern.notice] <6>[fffffe8004df0c80] TLE_INFO: Tape [VA0437], 10001510 shrunk. THL=0 Current allocated size 40960MB

Nov 26 20:58:30 stkvtl05 genunix: [ID 459543 kern.notice] <6>[fffffe8004df0c80] TLE_ERROR: Unable to reserve space for a new metadata segment at THL=0
Nov 26 20:58:30 stkvtl05 genunix: [ID 459543 kern.notice] <6>[fffffe8004df0c80] TLE_INFO: WRITE: Failed, Callback already done, next command will fail..

Nov 26 20:58:30 stkvtl05 genunix: [ID 459543 kern.notice] <6>[fffffe8004df0c80] TLE_INFO: Vtape 10001510 [VA0437].
Nov 26 20:58:30 stkvtl05 sanidmgr[673]: [ID 702911 daemon.notice] IPSTOR||0||E||0x00009cbc||Write block list failed

NOTE: Root cause not determined. Engineering could not determine the cause from the Xray logs, but since it was a specific time and a small number of tapes, the most probable cause is a temporary network glitch that caused timeouts within the processes, causing the vtape to be "stuck". And since this happened to tapes that had expired and were being relabeled, no data was lost.

Need vtape header info dumped and reviewed to find root cause.

Solution

Check to see if the virtual tapes in question are still mounted in virtual drives. If so, dismount them via the VTL Console (right click on drive where mounted and move to slot).

Since these virtual tapes are being relabeled, the data is already expired, so the quick solution is to manually delete and recreate the vtapes via the VTL console (refer to VTL Users Guide). Then have backup application inventory to pick up and use new tapes.

To have engineering look into root cause, dump vtape header(s) and escalate to engineering for review.
See doc id 1346834.1 for procedure on dumping vtape headers.

Note: When you get a Service Request on this issue, even though it seems like it is just simple write error and the solution mentioned in this document will fix it, it can be more than what it seems. Before following the steps above to troubleshoot the issue, it is always recommended to get the Xray output (Document 1185517.1)from the system, to further troubleshooting the issue or escalating to Falconstor, if needed.

Attachments

This solution has no attachment