I have the following environment:
1 – BE2012 CASO server (2008 R2) at our corporate office, dedupe as primary storage, tape secondary
3 – BE2012 MMS’s (2008 R2) at three plant locations, dedupe as primary storage, tape secondary
These servers are up to HF 201596, but do not have SP2 applied yet.
Backups from the plant locations are duplicated to the corporate dedupe folder. Some sets are also duplicated to tape at the plant location or at the corporate office for longer term storage.
We are getting an error while trying to restore data from the dedupe storage to one of our Windows 2003 SP2 servers (the server is virtual, Hyper-V, but being backed up individually using the Windows Agent). Only selected folders were being restored, and not to the original location, but to a different drive & folder on the same server. The job failed with:
Final error: 0xe00084ca - The data being read from the media is inconsistent.
Final error category: Backup Media Errors
For additional information regarding this error refer to link V-79-57344-33994
The job failed when it was loading the second “media” from the dedupe folder. I have tried restores from both the CASO and the MMS, with different sets from 2 different dedupe folders, and also to another 2003 SP2 server (also a virtual, but backed up individually with the Windows Agent). All of these backups were also duplicated to tape with software encryption, and they all fail with the exact same error when the second “media” is loaded.
To see if the duplicate to tape stage was causing a problem, I then ran a full backup of one of the virtual 2003 servers (using the Windows Agent) and set the duplicate to tape stage on hold. On a restore of selected folders again as before, it still failed with the same error when the second “media” was loaded.
I then tried a restore of data from one of our 2008 R2 physical servers; same scenario of restore selected content from dedupe to a different location on the server. I selected enough content so it would have to load another “media” from dedupe. It completed successfully.
I am going to run some more scenarios to see if I can narrow it down if this is only occurring on restores to virtual servers and then if only 2003, but wanted to get this out in case someone else has run across this, since I have wasted so much time already trying to figure out what is causing this.
I located
http://www.symantec.com/business/support/index?page=content&id=TECH152881
which sounds like a similar situation, but this was for 2010, so I didn’t know if the workaround registry change would apply to 2012, and I am not using encryption on the dedupe folder, only on the duplicate to tape (which I did try the same restore from the encrypted, duplicated tape, and it was successful).
The following is a section from SGMon log around the time when the second “media” is loaded on the dedupe:
BENGINE: [09/01/13 13:50:12] [4868] 2013-09-01T13:50:12.312 [loops] - RestoreBSDProcessor::TapeChange returning result = 0x00000000
BENGINE: [09/01/13 13:50:13] [4868] 2013-09-01T13:50:12.685 [loops] - stream data[DEFAULT] halted: INTERNAL ERROR
BENGINE: [09/01/13 13:50:13] [4868] 2013-09-01T13:50:12.685 [loops] - data halted: INTERNAL ERROR
BENGINE: [09/01/13 13:50:13] [4868] 2013-09-01T13:50:12.758 [loops] - notifyMoverHalted: Halt reason = HALT_CONNECT_CLOSED
BENGINE: [09/01/13 13:50:17] [4868] 2013-09-01T13:50:17.616 [tpfmt] - TF_FreeTapeBuffers: from 1 to 0 buffers
BENGINE: [09/01/13 13:50:18] [4868] 2013-09-01T13:50:17.766 [loops] - ****** deleteing *tapeServer - 0xaddf800
I have been waiting to apply SP2 to see the feedback, but going through the release notes, I don’t see anything related to this issue. Has anyone else run across this issue on restore?