See

http://mailman.eng.auburn.edu/pipermail/veritas-bu/2000-December/001845.html 

for 1 and 2.

Winkeler, Paul writes:
> All,
> 
> The SA's here are getting tired of being paged late at night because
> NetBackup encountered yet another 84 error.  Instead they feel it should be
> "straightforward" to take the information passed to the backup_exit_notify
> script and take automated action.  Here are the actions they would like to
> see:
> 1) On a "media" error, the volume causing the problem should either be kept
> from being used in any subsequent backup attempts.  This means:
>       freezing the volume if it is already allocated
>       moving it into a "bad tape pool" if it is not
> 2) Keeping track of the above failures, should a drive be involved more than
> once, DOWN the drive
> 3) If a stream fails, and it is part of a larger constellation of streams
> (we can determine this from the class name), kill off the other streams.
> 
> All these responses require more information than what the script is
> provided with.  Has anyone else come up with a technique to divine such
> facts as MediaID, Drive#, etc., by create use of the panoply of bp... tools?
> 
> Still amazed at NetBackup's lack of management interfaces I remain
> thoroughly humbled,
>  Paul
> 
> ----
> Paul Winkeler, Consultant
> +216-471-3795
> _______________________________________________
> Veritas-bu maillist  -  Veritas-bu AT mailman.eng.auburn DOT edu
> http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu

-- 
Larry Kingery 
                           Free the mallocs