After a server update from SLES11 SP2 to SP3 our nagios server was giving us a red light saying that one process on the updated server was not running anymore.
The check is done via NRPE using the command /usr/lib/nagios/plugins/check_procs -c 1: -w 1: -a “-Dapp=jportal”. Calling the command from the CLI on the updated server I get the expected result (1 process is running). If I use the old check_procs from SP2 the NRPE call is working again.
After some tests I’ve found that I get the expected result via NRPE as well using the command env COLUMNS=500 /usr/lib/nagios/plugins/check_procs -c 1: -w 1: -a “-Dapp=jportal”.
So I guess that the new check_procs has a bug and is not setting the COLUMNS variable big enough for its ps call to get the complete process output from ps.
An updated version of check_procs is much appreciated.