Hi all,

We're running NetBackup BusinesServer v.3.3 on a Sun Box with Solaris 2.6.
The box is connected via SCSI straight to a Qualstar TLS-4000 tape robot
with an AIT-2 drive.

We've been working with Veritas on fixing a number of issues regarding our
setup, but now we are stumped on this one. We keep getting an 84 error
sometimes when backups are run. A know it's not a problem with the tapes, as
they have been replaced with brand new ones, and the drive has been cleaned.
I'm thinking that this may be an SCSI issue, as in var/adm/messages, we keep
getting the following ouput daily:

Sep 17 11:02:45 buster unix: WARNING: /sbus@3,0/SUNW,fas@3,8800000/st@1,0
(st1):
Sep 17 11:02:45 buster unix: SCSI transport failed: reason
'unexpected_bus_free': giving up
Sep 17 11:03:58 buster unix: fas: 1.0: cdb=[ 0x3 0x0 0x0 0x0 0xc 0x0 ]
Sep 17 11:03:58 buster unix: WARNING: /sbus@3,0/SUNW,fas@3,8800000 (fas0):
Sep 17 11:03:58 buster unix: Connected command timeout for Target 1.0
Sep 17 11:03:58 buster unix: WARNING: /sbus@3,0/SUNW,fas@3,8800000/sd@3,0
(sd3):
Sep 17 11:03:58 buster unix: SCSI transport failed: reason 'reset':
retrying command
Sep 17 11:03:58 buster unix: WARNING: /sbus@3,0/SUNW,fas@3,8800000/sd@c,0
(sd11):
Sep 17 11:03:58 buster unix: SCSI transport failed: reason 'reset':
retrying command

Can anyone tell me if this is linked the the backup failures? The robot and
drive are the only devices linked to the SCSI cable. Prehaps a bad cable?
The cable is under the minimum length, and appears to have the correct
terminators. Veritas are stumped by this one, so i'm hoping someone here
will have some clues.


Thanks in advance,
--
--
Robert Chatfield