I have a DPM recovery job from tape failing, trying to recover a file server VM (1,2TB across 6 vhd files). First I tried to recover from tape only the .vhd that i'm looking for (around 350GB .vhd) but DPM fails miserably at the begining of the job with error ID 1001 (no further data, sorry). My only chances are recovering the full VM wich is spanned across 2 LTO4 tapes in the library. The job starts OK and I can see the first 4 vhd files of the VM in the destination path of the recovery job. The 5th .vhd (the one I need) is spanned to the 2nd tape and it's where DPM fails, not allowing me to recover the 5th and 6th vhd. I believe DPM cannot recovery data spanned across tapes with my library configuration.
I have a retention policy of 4 weeks in this Protection Group, I tried to recover from the last 3 TapeSets allways with the same error. I tried changing the BufferQueueSize in the registry to 1,2 and 3 from the default 10, with no results.
Error output:
The recovery jobs for Microsoft Hyper-V \Backup Using Child Partition Snapshot\<File Server VM> that started at quarta-feira, 20 de Março de 2013 14:53:12, with the destination of <DPM Server>, have completed. Most or all jobs failed to recover the requested data. (ID 3111).
DPM encountered a critical error while trying to parse the contents of the tape Hyper-V Cluster VM (Weekly)-00000011 (000028LU). This may be because the data on this tape is corrupt. (ID 24080 Details: Internal error code: 0x80990AA5)
My DPM Server configuration:
Server: Dell Poweredge 2950
Tape Library: Dell TL2000 SAS (IBM 3573 rebranded) with IBM Corporation driver 6.2.1.5 dated 13-01-2011 and firmwareA.50
OS: Windows Server 2008 R2 with SP1 (all latest patches installed)
DPM: DPM 2012 RTM (UR3 installed) supported on SQL 2008 R2
