Cleaning up metrics

[tonyleatwork]

Hi -

Is it possible to modify a service alert to prevent it from being in an alert state? Some alerts we would like to keep for the metrics but the options prevent me from skipping the -warning and -critical options. Basically can we override the output to have it be 'OK' all the time from the service definition.

Also is it possible to prevent alerts from going into an UNKNOWN state? We have some remote sites that are naturally getting timeouts no matter how we tune our network. While we're at it, can we clear up all the recorded "UNKNOWN" alerts so that it isn't counted against our metrics?

Regards,

Tony Le

[avandemore]

You can write a custom plugin to always exit with a return code of 0 to prevent any warnings/critical/unknown states. Or just write a wrapper for existing ones which does the same.