Hi gurus,

Environment:

NBU 4.5 DC MP3 on Solaris 8 with latest patches.
L700 (10xDLT8000) connected via 2 x Crossroads 4450
bridges.


Problem description:

When NBU server starts policy for backing up some data
(e.g. on three drives in parallel) some errors are produced
after couple of minutes. For the first minutes tapes are
mounted and after that fully utilized (average xfer 5MB/s).
When transfer appears errors are produced:


Feb 25 16:28:41 clasp tldd[25220]: [ID 832037 daemon.error] scsi command 
failed, may be timeout, scsi_pkt.us_reason = 7
Feb 25 16:28:42 clasp tldd[25225]: [ID 832037 daemon.error] scsi command 
failed, may be timeout, scsi_pkt.us_reason = 7
Feb 25 16:28:44 clasp tldd[25270]: [ID 832037 daemon.error] scsi command 
failed, may be timeout, scsi_pkt.us_reason = 7
Feb 25 16:28:45 clasp tldd[25284]: [ID 832037 daemon.error] scsi command 
failed, may be timeout, scsi_pkt.us_reason = 7
Feb 25 16:28:46 clasp tldd[25292]: [ID 832037 daemon.error] scsi command 
failed, may be timeout, scsi_pkt.us_reason = 7
Feb 25 16:28:47 clasp tldd[25302]: [ID 832037 daemon.error] scsi command 
failed, may be timeout, scsi_pkt.us_reason = 7


Sometimes job returns 0, sometimes 41 status code (Network connection 
timeout).
Suppose that, the reason of 41 status code is that master server can't 
receive
data from drives in specified period of time. I'v checked documentation for
scsi_pkt  (SCSI packet structure) especially for pkt_reason definitions 
and noticed
that reason number 7 is -> CMD_TIMEOUT - Command timed out


Questions:

Does anybody have ever seen these errors before ?
What is the reason of SCSI command timeout ?


regards,

Darek