I understand your point, however I'm looking at a larger scale: there are many plugins that can be used for multiple purposes. check_http is normally used to check the health of an http server/service, however I can write a simple http "application" that turn it into a remote monitoring tool. Now I may want to tell check_http what state I want to have timeouts to return because I already have monitoring on the server itself and I really care about what the service state is, not the http server that is running my check.
Same thing goes for SNMP. If I use snmp for controling devices, I may want to monitor and alert on its responsiveness. OTOH snmp checks that timeout or fail because there is a problem with the sub-agent may have to be handled differently. It's a pretty obvious one for NRPE, check_*sql, etc. too. I think giving more control to the user on the state returned for a timeout on most plugins where it makes sense will help you achieve your goal.. my point is that it's not limited to check_*_time. Regards, -- Thomas -- Reply to this email on GitHub: https://github.com/monitoring-plugins/monitoring-plugins/pull/1246#issuecomment-40756036
