Turns out that the problem I was facing with drives going down, and 
vmoprcmd showing no tapes in any drives while there were tapes, was 
because I killed atleast 15 active jobs going to about 8 drives. I 
initially thought maybe my drives were not mapped right, but they were.

If I kill a large number of jobs, things go haywire. Tapes are left in 
the drives, but NBU seems to lose track of them. Stopping/restarting 
daemons any number of times on any server does not help. Inventory says 
there's nothing to update. I end up IPL'ing the STK 9710.

Anyone else seeing this problem? Is it a Library problem?

HP 11.0, 3.4.1, Master with the lastest patch and also the patch for 
Java GUI.
10 drives connected to HP bridges.
4 media servers with SSO. All HPUX 11.0 with latest patches.

Thanks
Prasad.