Go to Master Server, or any server that can see all drives if in shared
environment,  Use /usr/openv/volmgr/bin/vmdareq, use perl, hopefully
every one of your drives has something common like lto1-drv-3 or
dlt-drv-4 or something that you can easily do a regular expression on,
but either way the output is nice because the drive names are all listed
right at the start like..
Lto1-drv-8 - AVAILABLE
Host1 UP
Host2 UP
Host3 SCAN_HOST UP 

and the status is indented so you can always do:


While (<IN>) {
if (/^(\w+)/) or something like that to get the drive name, build a hash
of drive names to hosts and then just check for UP, if its not up, then
you can issue your drive up command, but a lot of times that doesn't
work because the drive either needs to be...

Cleaned, have a tape manually removed, powered on/off, or have a service
call replaced.  

Finally, we have many drives that GO Down or PEND-TLD and if we just
reset the fiber channel to SCSI bridge they connect through, then they
come right back up.  IF you have SCSI drives, but connect them over a
SAN via Fiber to SCSI bridges, this is usually a good thing to try.  You
won't hurt anything, just make sure no backups are going on on any other
drives that share the same fiber bridge when you do this.  I think (not
sure technically what happens) but these bridges get out of sync between
fiber/scsi commands and then bam, the drive can't talk to the other side
and netbackup downs it...