I’ve read the doc on setting up distributed nagios a couple times. The question I have is does the main server HAVE to be set to not do active checking?
We setup a Nagios environment to monitor all of the servers we have which works great. However we wanted to build some small nagios servers to put at some of our clients locations that could be set to monitor their onsite servers and passively send the results back to us. If I have to have passive only checks on the main server then I’d have to restructure everything which I’d rather not do if I can avoid it.
Just hoping to get some clerification.
Thanks in advance.
–Aric
You can mix and match as you like. I use passive service checks at the moment only for SNMP trap processing; the rest are active checks.
You can also install nrpe on your monitored nodes and send passive checks back to the main Nagios server without needing an intermediate full-blown Nagios server (but installing nrpe on each monitored node might turn out to be more hassle than its worth).
That is not passive, but active. The check in your suggestion would be initiated by the nagios central server. The remote host, would then run the check and send the data to via nrpe to central server client connection. Again, that is considered an active check.
If you are setting up a service check to be performed on the remote host, then you surely don’t want the same check to be performed on the nagios central server. So, for that one service check, you should disable active checks. Make sense? If you don’t, then one minute the results will be from the remote system, then another minute later, they are from the central server.