We have definitely seen similar behavior with HP skinned Crossroad FC/SCSI 
bridge.  At the beginning of our implementation it was devastating, causing 
backup to crash approx. every three days.  A firmware upgrade corrected what 
was identified as a buffer issue.  Then we were in a similar position as 
described with failures every other week or so.  At that point no additional 
firmware upgrades were available, so scripted to have our three bridges reboot 
nightly prior to our backups starting.  This seem to make our situation much 
more stable. We have upgraded firmware on the bridges since then, but have been 
reluctant to stop the reboots.

Patrick Wilhelm
Davidson College - ITS
704-894-2569
pawilhelm AT davidson DOT edu


-----Original Message-----
From: rene_casalme AT baxter DOT com [mailto:rene_casalme AT baxter DOT com]
Sent: Friday, September 19, 2003 8:52 PM
To: veritas-bu AT mailman.eng.auburn DOT edu;
veritas-bu-admin AT mailman.eng.auburn DOT edu
Subject: [Veritas-bu] SAN Bridge Problem and NetBackup


Hello all,

I apologize if this is off-topic but I don't know of a better group to ask
this question.  If you know of another list, please let me so I could post
there instead.

We are currently running NetBackup DC 4.5 FP3 and Compaq MSL tape library
connected via a Compaq StorageWorks Modular Data Router (bridge).  Every 10
to 15 days or so, the NetBackup master experiences scsi command errors as
shown in the /var/adm/messages file.  It is specifically the tldcd and the
HBA card that are showing these errors.  For the meantime, I reboot the
bridge and 95% of the time they all see each other.  The other 5% we
restart NetBackup and/or server.

I get this same problem with another bridge (StorageTek) in another SAN and
NetBackup master setup.

Anybody else sees this problem?  Any pointers will be much appreciated.

Rene



_______________________________________________
Veritas-bu maillist  -  Veritas-bu AT mailman.eng.auburn DOT edu
http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu