Hey folks. Sort of a philosophical question, here:
Currently we perform a bunch of resource checks via nrpe. The nrpe daemons are configured specify when a resource is OK, in warning, or critical state. My question: How can I get values from the daemon, and have the server figure out when something is not OK?
Has this been considered, or completely out of the nagios/opsview model? I've used nagios for over a decade, but don't remember coming across something like this, but now as I consider a more programatic rollout of configurations and such, it seems like having the server know what's good/bad makes more sense.
Examples: CPU temperatures will vary from one end of a datacenter to another. A SSD in a server will have ideas of "high" bandwidth compared to a spinning disk. I could create nrpe check scripts with the smarts to figure this out, but just seems like having it server-side might be a better idea.