CA Arcserve Backup R12 SP2 - Scanning Tapes Problem

backward compatibility is 100%

check ARCserve home directory\log\tape.log at this time 16/10/2009 1:50:18 PM

It will have something like this
2009/10/16 14:51:56 [1038] =>ABSL:3050 [REQUEST SENSE   ] 03 00 00 00 40 00 00 00 00 00
2009/10/16 14:51:56 [1038]             <TEST UNIT READY     >, Sense Data as Follows:
2009/10/16 14:51:56 [1038]          SENSE ABSL:3050      70 00 06 00 00 00 00 10 00 00 00 00 28
2009/10/16 14:51:56 [1038]       EX SENSE ABSL:3050      00 00 00 94 09 00 00 00 00 00 00 00 00 00 00 00
2009/10/16 14:51:56 [1038]          SENSE ABSL:3050      Unit Attention [06]
2009/10/16 14:51:56 [1038]       EX SENSE ABSL:3050      Not Ready to Ready Transition(Medium May Have Changed) [28, 00]


details in your case will be completely different, but the important part is Request Sense is the SCSI command sent by the ARCserve Tape Engine to the tape device requesting the devices sense codes which have details on the command that failed. Following lines are those sense codes from the drive.

That said chances are that based off message in the activity log the command that failed was a READ command and the reason is the drive, even after many retries could not read tape.
 
As a test of the drive try a restore from a brand new brand new backup, if you are lucky it will fail and you have a bad drive. Lucky??? yeah I'

sorry hit submit by mistake.

yeah I'm figuring you would rather have a drive problem than for all your data tapes to be bad, data is usually worth a lot more than the drive.

by the way what brand tapes are used?

Hi Everyone,

Thank you all for the responses so far!

I have enabled verbose logging on the Tape Engine as suggested (prashant9885) and found the following in C:\Program Files\CA\ARCserve Backup\LOG\CADBLog.log:

2009/10/19  7:29:45:366 - 1860,7836: [SQL] SQLDBAddAstperrRec(rc=SQL Error State:22001, Native Error Code: 1FD8, ODBC Error: [Microsoft][SQL Native Client][SQL Server]String or binary data would be truncated.SQL Error State:01000, Native Error Code: E25, ODBC Error: [Microsoft][SQL Native Client][SQL Server]The statement has been terminated.)

2009/10/19  7:51:50:360 - 1860,9224: [SQL] SQLDBAddAstperrRec(rc=SQL Error State:22001, Native Error Code: 1FD8, ODBC Error: [Microsoft][SQL Native Client][SQL Server]String or binary data would be truncated.SQL Error State:01000, Native Error Code: E25, ODBC Error: [Microsoft][SQL Native Client][SQL Server]The statement has been terminated.)

**********************************************************************************
prashant9885:

We have it connected via SCSI.

You mentioned disabling the drivers for the Tape Loader and the Tape Drive. In Device Manager we have the following:

Medium Changers:
Hewlett Packard MSL G3 Series Autoloader (x32 based 2003)

Tape Drives:
Hewlett Packard LTO Ultrium-2 drive

We also have the Removable Storage Service running and set to Automatic.

So simply disabling the devices in Device Manager will be ok? I have been told that Disabling the Removable Storage Service should be done as well. Please let me know :)

**********************************************************************************
dovidmichel:

I checked the log and this isn't the best thing to see on a Monday morning... :(

Sounds like Arcserve can't connect to the Database... (Possibly to verify the tape?)

[10/16 13:41:05 2134  1 TK        9367   9] Default Tape Block Size for Device [2] is [65536]
[10/16 13:41:05 2134  1 TK        9367   9] Current Tape Block Size for Device [2] is [65536]
[10/16 13:41:05 2134  1 TK        9367   9] Default Tape Block Size for Device [2] is [65536]
[10/16 13:41:05 2134  1 TK        9367   9] Current Tape Block Size for Device [2] is [65536]
[10/16 13:50:18 20b4  1 TK        9367   9] =>ABSL:5040 Command:[SPACE               ] <Drive>
[10/16 13:50:18 20b4  1 TK        9367   9]   ABSL:5040 Sense Key status: Media Error [03]
[10/16 13:50:18 20b4  1 TK        9367   9]   ABSL:5040 Additional sense: Recorded Entity Not Found [14, 00]
[10/16 13:50:18 2138  1 TK        9367   9]         DRV:2 [HP      Ultrium 2-SCSI  S54W] recovered [0] error for [reading.].
[10/16 13:50:34 2138  1 TK        9367   9] DB: ReConnect to DB failed, set DB state flag as abnormal
[10/16 13:50:34 2138  1 TK        9367   9] ClientDisconnectFromTape(): set tape OFF -LINE
[10/16 13:50:34 2138  1 TK        9367   9] !!!!! Bad tape in Device[2] was from Slot[11].
[10/16 13:50:34 2138  1 TK        9367   9] !!!!! We now try to return this tape back to its slot, and mark it as UNREADABLE MEDIA.
[10/16 13:51:14 2138  1 TK        9367   9] !!!!! Tape in  Library[3][SN:DEC06341HD] Slot[11] has been marked as EMPTY!!

Regards,
Stefano

DovidMichael,

I tried to scan the daily backup that we just ran last night and the scan worked fine, all sessions came up and there was no read errors. It is looking more and more to be those old sets of tapes... which is a real problem!

The loader and drive seem to be fine. I believe the SQL database error could have been the fact that it couldn't scan the tapes at all.
I have requested a newer set of monthly tapes from our onsite people and will see if we can finally find some decent tapes. Thanks for the answer. :)