There is nothing in the log close to the time the TSM server reported these errors, but there are errors at an earlier and late time. See below:
2008-05-23 01:47:26: NDMP: 4: cifs_ea new count: 223660947, delete count: 223660904
2008-05-23 01:47:38: DRIVERS: 4: FCDMTL 2 [9.0.2] Initiator check condition: loopID = 0., SK = 03, ASC/Q = 3100
2008-05-23 01:47:38: CAM: 3: I/O Error: c96t0l0: CamStatus 0x84 ScsiStatus 0x02 Sense 0x03/0x31/0x00
2008-05-23 01:47:38: CAM: 3: camFlags 0xc1 Addr 0xffffffff Len 0x0
2008-05-23 01:47:38: CAM: 3: cdb: 10 00 00 00 01 00 00 00 00 00 00 00
| SCSI ERR ON #9600 camstat 0x84 bstat DeviceError
| scsi stat.skey.ascq 0x02.03.3100 Check condition
| skey [ ] Medium error
|____________________ ascq Medium format corrupted




2008-05-23 05:09:21: DRIVERS: 4: FCDMTL 2 [9.0.2] Status buffer overrun: xid 22d, msg 443fe00, address id 17a00
2008-05-23 05:09:21: DRIVERS: 4: FCDMTL 2 [9.0.2] Initiator check condition: loopID = 0., SK = 00, ASC/Q = 0001
2008-05-23 05:09:23: CFS: 4: fs 0x11f type = dhfs being unmounted. Waiting for quiesce ...
2008-05-23 05:09:23: CFS: 4: fs 0x11f type = dhfs unmounted
2008-05-23 05:09:23: VRPL: 4: Playback: v:359, Delta:6426, g:6487, ad:15991808, nDelta:1.
2008-05-23 05:09:24: DRIVERS: 4: FCDMTL 2 [9.0.2] Status buffer overrun: xid 28, msg 443ff80, address id 11a00
2008-05-23 05:09:24: DRIVERS: 4: FCDMTL 2 [9.0.2] Initiator check condition: loopID = 1., SK = 00, ASC/Q = 0001
2008-05-23 05:10:06: IP: 4: 10.168.17.177:138 - 10.3.49.101:138 address block released
2008-05-23 05:10:12: DRIVERS: 4: FCDMTL 2 [9.0.2] Initiator check condition: loopID = 0., SK = 02, ASC/Q = 0401
2008-05-23 05:10:22: DRIVERS: 4: last message repeated 1 times
2008-05-23 05:10:22: DRIVERS: 4: FCDMTL 2 [9.0.2] Suppressed all but 2 out of 10 identical log entries
2008-05-23 05:10:22: DRIVERS: 4: FCDMTL 2 [9.0.2] Initiator check condition: loopID = 0., SK = 06, ASC/Q = 2800
2008-05-23 05:10:25: DRIVERS: 4: FCDMTL 2 [9.0.2] Status buffer overrun: xid 1cc, msg 4421420, address id 17a00
2008-05-23 05:10:25: DRIVERS: 4: FCDMTL 2 [9.0.2] Initiator check condition: loopID = 0., SK = 00, ASC/Q = 0001
2008-05-23 05:10:29: DRIVERS: 4: FCDMTL 2 [9.0.2] Initiator check condition: loopID = 1., SK = 02, ASC/Q = 0401
2008-05-23 05:10:39: DRIVERS: 4: last message repeated 1 times
2008-05-23 05:10:39: DRIVERS: 4: FCDMTL 2 [9.0.2] Suppressed all but 2 out of 10 identical log entries
2008-05-23 05:10:39: DRIVERS: 4: FCDMTL 2 [9.0.2] Initiator check condition: loopID = 1., SK = 06, ASC/Q = 2800
2008-05-23 05:10:42: DRIVERS: 4: FCDMTL 2 [9.0.2] Status buffer overrun: xid 170, msg 4422140, address id 11a00
2008-05-23 05:10:42: DRIVERS: 4: FCDMTL 2 [9.0.2] Initiator check condition: loopID = 1., SK = 00, ASC/Q = 0001

what does the Celerra side say ?

take a look at "server_log server_x" for around the time of the failure

is it being backed up to physical library or VTL ? Any problems with the library itself ?