Question : Problem: backup failing with SQLVDI error

I have two active active two node clusters. Even number nodes 2 and 4 have a terabyte database while the other nodes have databases in the hundreds of gigbytes. About a month ago the larger databases started failing without any real indication in all the logs. I can send anywhere from 100 gigabytes of data to 1 terabyte of data but yet it still fails before completing. All other databases some on the same node as the 1 terabytes complete although they may be 150-200 gigbytes.

The only real indication or error is the following:

SQLVDI: Loc=SignalAbort. Desc=Client initiates abort. ErrorCode=(0). Process=7876. Thread=9868. Client. Instance=server4. VD=Global\TDPSQL-00001EC4-0000_SQLVDIMemoryName_0.

I already applied the vdi patch to SQL 2005 sp2.

Could it be the TSM server, possibly, could it be the VTL, possibly, could it be the TSM code level, its all possible. However, the only issue or error I see if SQLVDI.

The backup application sends up to a terabyte of data without an error and then boom, it fails.

None of this makes sense..... The databases are higly utilized. A month ago I was able to witness better speeds then today. My issue is nothing I can prove.

Transaction logs are taken every 15 minutes to local disk. Rebuild indexes occurs and in addition a purge job exists. With so many components and teams, and the fact that its production i dont have the freedom to test what I want.

I am only searching for idea....

Please offer....

Answer : Problem: backup failing with SQLVDI error

Yes I have applied that patch back about 30 days ago and the issue persists. I also applied every TSM setting possible. I think the best resolution at this point is to award partial points and move on because all our part of the solution eventhough it has yet to be resolved.