Well I finally have my setup figured out, and I'm about ready to deploy my distributed systems, but I have a questions (well more of a mental exercise).
I would like to be able to trigger service/host checks on one of the remote nodes from the master server.
I would like the service on the main host to show up as a warning if the service goes stale, then contact a forced check on remote server. Could I do this with nrpe?
What processing logic could I use here to get this process:
- Service is stale
- Set service to 'warning'
- Trigger the forced check plugin
- Plugin writes to remote servers command file to force the check
- If the plugin successfully writes to the file, go to hard warning state to wait for passive check to come in
- If the plugin fails, retry on set retry_interval
- If plugin fails after max_attempts, go to hard critical and force an active host check on the remote server (which can be directly pinged from the central server)