I need to check if a backup process has hung. That is, the backups on the system being checked normally finish in under 2 hours, so I would like to find any backup process that is older than 2 hours.

The nagios plugin check_procs lists the option of elapsed time, but does not seem to work in my environment. I have compiled plugins version 1.4.12 on AIX 5.3.

Anyone have any suggestions?

Hello and welcome to LinuxQuestions,

What's the exact command you are running, i.e. with what parameters are you executing the check_procs plugin?

Kind regards,

Eric

Here is the command:

/usr/local/nagios/libexec/check_procs -m ELAPSED 120 -C tsm_db2_backup

This does find a process (when it exists) but no matter how much elapsed time this process has, this command returns

ELAPSED OK: 1 process with the command name "tsm_db2_backup"

Note: I have tried to check commands like init and sshd, but still, no success.

In doing this follow up, I see that check_procs version 1.4.12 for AIX does not have the option metric=ELAPSED. (The Linux version does) And, AIX is where I need this option.

In the end, what I need is a way to find a runaway process. If you know of other checks for this, please share.

Thanks for taking a look at this problem. All suggestions are welcome.

Hello,

I noticed the 'tsm' in the command name you used so I imagine that you're checking up on Tivoli Storage Manager backups. Have a look at this script:
Check TSM on AIX. Maybe it'll give you better results.
From the same site there's also this one.

Kind regards,

Eric

I have found a shell script that does what I need

http://exchange.nagios.org/directory...untime/details

check_proc_time.sh works under bash or ksh (change first line for proper shell), so I get the results I need under AIX.

check_proc_time.sh -p tsm_db2_backup -w 3600 -c 7200

gives a WARNING if tsm_db2_backup runs for 1 hour, and CRITICAL if it is running for 2 hours - and OK if process NOT running. Just what I needed.

Thank you all for taking a look at my query.

Hi,

That's great news. Glad to hear you found a solution and thanks for posting the link here. If you consider your problem/question solved please mark this thread as solved using the Thread tools.

Kind regards,

Eric