The new check_load settings are too low

[Joi Owen]

The new default check_load definition in /etc/nagios/nrpe.cfg is way too low for real-life. It seems the command definition changed between Ubuntu 16.04 and 18.04 to use a -r option, but I'm finding this simply causes false alarms the moment a nagios client is configured to respond to check_load with nrpe. I'm having to go back through all the 18.04 clients I monitor and manually override the default values with sane values placed in nrpe_local.cfg instead.

I don't know if the -r option simply isn't working as expected, or if the default values now used are just too low. A newly-built 18.04 is running BitDefender that that alone is causing non-stop warnings and criticals from nagios now, because now it's warning at 15% of available CPU.

Old version:

/etc/nagios/nrpe.cfg:command[check_load]=/usr/lib/nagios/plugins/check_load -w 15,10,5 -c 30,25,20

New version:

/etc/nagios/nrpe.cfg:command[check_load]=/usr/lib/nagios/plugins/check_load -r -w .15,.10,.05 -c .30,.25,.20

You want a host to have its cpu load up to match the number of cores it has, so you'd want these up around 1, at minimum, for 100% busy, not 15% of busy. That's just silly.

[Robie Basak (racb)]

Thank you for taking the time to report this bug and helping to make Ubuntu better.

It looks like this default originates from upstream's sample configuration here: https://github.com/NagiosEnterprises/nrpe/blob/master/sample-config/nrpe.cfg.in#L299, and https://github.com/NagiosEnterprises/nrpe/issues/171 and https://github.com/NagiosEnterprises/nrpe/pull/205 references it. Will that fix your problem? If so, I suggest discussion upstream to try and get that pull request landed first.

Importance: Medium since a workaround is available (local override in /etc/nagios/nrpe_local.cfg as described in the bug description).