1

I'd like to make order in our Nagios, in the check_load service.

There is no hardware baseline in my company and each server has it's own different specs, which means that the check_load script doesn't give the correct results on each server.

I'm looking for a dynamic Nagios check which checks how many processors/cores/threads exist in a server and edits the check_load command accordingly based on the numbers found.

Another idea I have, is to use facter to get that information and make Puppet edit the local nrpe.cfg template based on the findings of facter to match best practice load thresholds.

Do you know if such a check exists?

1 Answer 1

2

From check_load --help:

 -r, --percpu
    Divide the load averages by the number of CPUs (when possible)

This isn't guaranteed to work on every OS... it uses sysconf to query the number of CPUs (_SC_NPROCESSORS_ONLN or _SC_NPROCESSORS_CONF). I seem to recall it not reporting the correct values on AIX. Not a problem on Linux though.

The oldest version I have installed anywhere is 1.4.15, and the -r option is available in that version.

2
  • Thanks, I'll check that. Do you know if it has any default values to check against?
    – Itai Ganot
    Feb 16, 2015 at 16:29
  • It just toggles "divide by the number of CPUs (cores)" on or off. I guess you could say it defaults to "off", or "divide by 1".
    – Keith
    Feb 16, 2015 at 16:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.