Perhaps you've solved your problem by now, but I was receiving 84 errors
on our LTO drives that could not be isolated.  Veritas indicated that an
84 error meaning that communication with the drive was interrupted in
some way, and that it happened below the application level.  After
trying various firmware upgrades on drives, HBAs, and FC/SCSI bridges,
we finally swapped out our Brocade switch and the problem went away.
For your situation it could be any link in your FC path.  I would be
interested to know what you find.


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



ferbraga AT vento.com DOT br ferbraga AT vento.com DOT br 
Thu, 3 Jan 2002 18:20:35 +0000 

Previous message: [Veritas-bu] Incredibly slow backups of NFS mounted
partitions from NetApps. 
Next message: [Veritas-bu] Restores using the Java GUI 
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] 

------------------------------------------------------------------------
--------

I have sometimes a failed backup with error 84. It occurs after NBU have
done a lot of work in this class (more than 500GB). So, I don't suspect
drive error, drive configuration errors, or media type errors, like
Troubleshoot
Guide suggests. The medias are all new, and the drive is clean. This
class
use a storage unit with two drives.
My master server is a Solaris8, Netbackup 3.4 with patch 34_2.
The media server, where the error occurs, is a Tru64 5.1 with the same
patch
level. The tape drive is a 9840FC. In this media server are another
backups
running pretty well.