Here's a variant of my old drive monitor script.  Just run it on your
master server (only) via cron as often as you want to sample.  Perhaps
every 15 min.
 
It'll actively lookup your media servers, check them, and report drives
down on that media server.  It keeps a history file so it'll only report
a drive once unless it's cleared then reoccurs.  Change the MAILADDR
variable for your environment.

$>cat drivemon 
#!/bin/ksh
 
PATH=$PATH:/usr/openv/volmgr/bin:/usr/openv/netbackup/bin/admincmd
 
MAILADDR=Netbackup.Support
 
# run-to-run history
HIST=/tmp/`basename $0`.hist
[ ! -f $HIST ] && touch $HIST
 
# temp file to gather vmoprcmd output
TMP=/tmp/`basename $0`.tmp
[ -f $TMP ]    && rm -f $TMP
 
#Log management
LOG=/usr/openv/netbackup/logs/`basename $0`.log
[ -f ${LOG}.5 ] && mv ${LOG}.5 ${LOG}.6
[ -f ${LOG}.4 ] && mv ${LOG}.4 ${LOG}.5
[ -f ${LOG}.3 ] && mv ${LOG}.3 ${LOG}.4
[ -f ${LOG}.2 ] && mv ${LOG}.2 ${LOG}.3
[ -f ${LOG}.1 ] && mv ${LOG}.1 ${LOG}.2
[ -f ${LOG}.0 ] && mv ${LOG}.0 ${LOG}.1
[ -f $LOG     ] && mv $LOG     ${LOG}.0
exec >$LOG 2>&1
echo "## Started: `date`"
 
for m in `bpstulist|awk '$5>=0 && $7>0 {print $3}'|sort|uniq`
do
  echo "## Gathering $m."
  vmoprcmd -h $m -xd ds|awk '$4~/DOWN/ {print "'$m' "$2}'|tee -a $TMP
done
 
while read mserver drive
do
  if [ `grep -c "^${mserver} ${drive}$" $HIST` -eq 0 ]
  then
    echo "## Drive $drive down on $mserver: Newly discovered."
    echo "Drive $drive on $mserver is down: `date`." | \
      mailx -s "NB Wrn: $mserver:$drive down" $MAILADDR
  else
    #Just debug info for logfile
    echo "## Drive $drive down on $mserver: Already Reported."
  fi
done <$TMP
 
mv -f $TMP $HIST
echo "## Done: `date`"
 
exit

________________________________

From: veritas-bu-bounces at mailman.eng.auburn.edu
[mailto:veritas-bu-bounces at mailman.eng.auburn.edu] On Behalf Of Jeff
Lightner
Sent: Friday, August 25, 2006 9:56 AM
To: veritas-bu at mailman.eng.auburn.edu
Subject: [Veritas-bu] queued job never going away



In the activity monitor on a few occasions we've seen jobs staying
"queued" forever.  Is there a way to get rid of such queued jobs?  

Canceling them or deleting them doesn't seem to help.

It appears they occur when a client initiated action occurs and goes
away on the client.  It somehow leaves it behind on the master.  This
has been seen with RMAN backups and MS-SQL restores.

The queued job doesn't prevent later client initiated operations from
running so its display is more of a nuisance than a real problem.
Restarting NBU does make it go away but due to the number of jobs we
have this is something we usually only do on a Friday when there are
issues occurring.  Has someone else has seen this and figured out a way
to get rid of the queued jobs short of restarting NBU?

Jeffrey C. Lightner

Unix Systems Administrator

DS Waters of America, LP

678-486-3516