Restoring the scripts section from a NVDB backup fails (4037307)

Restoring the scripts section from a NVDB backup fails

From the Binary log we could confirm the following messages and errors for NVDB scripts restore job:

1] Information   2014/01/20 16:10:49   6012   Data Plugin   prao_sib_fr   Restoring script file: 'mediadetails.PK0096' size: '0'
2] Error   2014/01/20 16:10:49   6012   Jobs   prao_sib_fr   Job manager lost message channel unexpectedly
3] Error   2014/01/20 16:10:49   6012   Jobs   prao_sib_fr   Job Status: Restore Failed

From the NVDB Trace (nvdbplugin1390234246):
2   DBRESTOR   ???   27737   92   0   161049085564   Converted '/opt2/netvault/scripts/mediadetails.PK0096' to '/opt2/netvault/scripts/mediadetails.PK0096'
2   DBSCRIPT   ???   27737   37   0   161049085572   Restoring scripts file: '/opt2/netvault/scripts/mediadetails.PK0096' size: '0'

Job Manager Trace (nvjobmgr6517):
2   JOBMGR   ???   27732   62   0   161049012535   Plugin is now in state Reading from media
0   NET   ???   27732   84   0   161049086110   Connection down
2   MESSAGE   ???   27732   1010   0   161049086218   Destination (null):3
2   MESSAGE   ???   27732   1011   0   161049086223   Source (null):3
0   JOBMGR   ???   27732   447   0   161049086410   Plugin lost message channel unexpectedly

Network Manager Trace (nvnmgr):
0   NETLOOP   ???   26158   193   0   161048579012   Lost connection connecting to 'PCCLT0023', state 1 Disconnect TRUE
0   NETLOOP   ???   26158   132   0   161048581643   Inform communication manager that connection to PCCLT0023 is down

We could confirm various scripts were being restored from the Binary log, but when the: 'mediadetails.PK0096' script was reached, the size was '0' and failed at that point.
The rest of the restored scripts had various different sizes.
The 'mediadetails.PK0096' script was on ‘0’ as had corrupted (the NetVault Server crashed earlier due to faulty HDD sectors).
So when the restore job had reached that script, the restore job couldn’t read it, therefore the message channel broke and the restore job failed.

The issue was resolved by restoring a non-corrupt scripts section of the NVDB.

Tip: Once all the items that you needed from a good NVDB backup have been restored; perform a NVDB backup so you have a more recent and up to date NVDB backup.
Generally we do advise that you should perform a two phase backup i.e. phase 1 to a VTL, SmartDisk or a DR4100/6000 device and then a phase 2 to a tape library, so that the tapes can be brought off-site or safe keeping. This way if one of the backup devices fail, you can restore from the other device.