Printer Monitoring Script - to report on Status - down, unresponsive Print Server, etc...

[Alzhy]

Because of an architectural brouhaha - our UNIX servers now hosts thousands of printers. The way we've set up our printers is we use the Windows NT/2K servers and their defined queues (Print Services for UNIX) -- so all queuing for Windows and UNIX is done on the Windows side of the equation.

Sometimes, the Windows NT server serving the printers would go down or its UNIX printing service (LPD) would go down or it's DNS domain changes, etc...

Does anyone have a script created to inventory and report on the status of all the Printers defined on UNIX?

Thanks...

[Mark Grant]

Nelson,

I am missing something here obviously but "lpstat -t" does a reasonably good job!

[Abdul Rahiman]

Nelson,
I was wondering the same too..
# lpstat -p does a good job for me.
If you want to know abt. only problem printers, you may grep the output like this.
# lpstat -p | grep -v -E "idle|fence|printing"

If you want to go further, you could also ping the remote lpd host's ip address to see if they respond via a script like this.

cd /etc/lp/interface/
ls ps* | while read lp
do
echo $lp
lpstat -p$lp | grep -v -E "idle|fence|printing" >> /var/adm/printer_stat.log
host=`grep "PERIPH=" $lp |grep -v MODEL| awk -F"=" '{print $2}'`
ping $host -n 2 >> /var/adm/printer_stat.log
done

HTH,
Abdul.

[Bill Hassell]

lpstat -t is always a problem. It does a query to all remote printers and it takes about 30 seconds to timeout for each printer on a dead server. The lpstat -p approach is better way and won't hang. However, here is a script that might solve all your print server problems:

rm -rf WindowsNT
install Linux
configure remote-printers
retrain WindowsAdmins

(sorry, I couldn't resist)

[Joseph Loo]

hi,

the status of the network printer needs only the lpstat command.

# lpstat -o
tells u the print currently in queue.
# lpstat -t
tells u if the printers are enable.

i frequently use the above 2 options for lpstat and this command should suffice if u need a script as well.

regards.

[Alzhy]

Thanks for your suggestions so far. As I've mentioned, each of our UNIX servers is configured to have thousands of printers that are defined on Windows Hosts (for centralized queue administration, etc..). I need a report each early AM to summarize what printers have problems. I have the following already but it takes about 7 hours to finish:

for i in /etc/lp/interface/*;do
echo Checking `basename $i`
echo =========================================================================

PRT=`basename $i`
lpstat -o${PRT} 2>&1 |egrep "unknown|waiting|down|bus">/dev/null

if [ $? -ne 0 ];then
echo $PRT is down|tee -a DOWN_PRINTERS
fi

done

[Sherin T. George]

Replacing
lpstat -o${PRT} 2>&1 |egrep "unknown|waiting|down|bus">/dev/null

with

rlpstat -d${PRT} -u 2>&1 |egrep "unknown|waiting|down|bus">/dev/null

might give a better performance.

[Sherin T. George]

One more point,
if you are interested only in remote printers you can query the directory /var/spool/lp/sinterface instead of /var/spool/lp/interface.