My current test script does nothing except print the severity on STDOUT and exit with the exit code I specify. When I have it exit with a 0, the icon turns green. If it exits with a 1, the icon turns red. All the others turn the icon to a circle with a question mark in it.
The past post just explained that a item is always either up or down. There were levels of up (ok, warning, alarm and critical). Thanks, Rich ---- Original message ---- >Date: Thu, 21 Jun 2007 21:56:16 -0400 >From: "Myers, Jon W" <[EMAIL PROTECTED]> >Subject: RE: [IM-Talk] event severities with a custom command line probe >To: "InterMapper Discussion" <[email protected]> > >Its up to the script you're actually running to exit with the proper codes. >For testing, change your script to just do an exit(4) and see if IM sees it >as a warning. >Depending on the script, you might have to make the timeout a little higher to >make sure it runs, the default is 3 seconds. My commandline probe (which runs >a CLI php script, which in tern runs digitemp to check the temps on 3 or 4 >sensors via 1wire bus) takes about 3 or 4 seconds to run. I believe IM's >default timeout is 3 seconds. > >I have no idea of your past messages, as I'm at home, and cannot lookup what >you posted a few weeks ago, but since noone has answered in about 6 hours, I >figured I'd toss out a small suggestion. > > >-----Original Message----- >From: [email protected] on behalf of [EMAIL PROTECTED] >Sent: Thu 6/21/2007 4:29 PM >To: InterMapper-Talk >Subject: [IM-Talk] event severities with a custom command line probe > >List members, > >About two weeks ago Bill Fisher answered my question about the different event >severities. Based on the way I understood his response I had our IM-pro >create a custom command line probe based on the probe provided by Dartware. > >We replaced the exit code definitions with an expanded section. The goal is >to allow separation between the network/OS caused errors and the errors >reported about the application. I'm not part of the network or OS teams and >do not need to be alerted about things I can not fix. > >This is the exit code section in our probe: ><command-exit> >critical:${EXIT_CODE}=6 >alarm:${EXIT_CODE}=5 >warning:${EXIT_CODE}=4 >okay:${EXIT_CODE}=3 >down:${EXIT_CODE}=2 >down:${EXIT_CODE}=1 >up:${EXIT_CODE}=0 ></command-exit> > >It is not working as expected. The only exit codes it recognizes is 0 and 1. >Everything else causes a ? icon. We tried removing the double definition of >down without any improvement. > >Is it possible to have the command line probe run a nagios type script, via >nrpe, on a remote system and understand severities beyond up and down? > >Thanks, >Rich >[EMAIL PROTECTED] >____________________________________________________________________ >List archives: >http://www.mail-archive.com/intermapper-talk%40list.dartware.com/ >To unsubscribe: send email to: [EMAIL PROTECTED] > > >________________ >TNEF25.rtf (3k bytes) ____________________________________________________________________ List archives: http://www.mail-archive.com/intermapper-talk%40list.dartware.com/ To unsubscribe: send email to: [EMAIL PROTECTED]
