Nagios Support Forum

On my XI demo, I'd like to retain and graph performace data from services, like memory usage, load, etc.

Where and what do I need to set in the UI, as it doesn't seem to do this by default?

I see "retain_state_information=1" in nagios.cfg by default.
I see "retain status information", "retain non-status information", and "Process perf data" in the Service Management Settings, all of which might be involved.

Would I have to go to each service on each host and change the Settings?

And if you can point me to any place in the XI docs where this is covered, I'd appreciate it, as I'm still learning my way around the XI docs. Maybe it's covered in the docs at pnp4nagios.org?

Thanks....Lyle Ryan, Cubic Corp.

A follow-on to my original post:

For every service on one monitored host, I checked the 3 Service config options mentioned above. Just a guess, so trying it.

Now the services show as "Unknown" status and an info field of "Check_nrpe: received 0 bytes from daemon. Check the remote server logs".

Our regular Nagios Core 3.0.3 is monitoring these nrpe services just fine. But I did notice that /var/adm/messages on the Solaris remote host is complaining that "request contained command arguments!" and "client request was invalid, bailing out".

So I'm guessing I have to set the nrpe.cfg on the remote host to include "dont_blame_nrpe=1". But I'm not sure how that will affect our 3.0.3 monitor, or why our 3.0.3 has been happy without passing parameters. Maybe the parameters are set in the remote plugins? ugh.

What should I do to get this running? I need to show off some performance graphs to justify the purchase.

Thanks....Lyle

In our 3.0.3 production Nagios, turns out that the remote host has an nrpe.cfg that defines the parameters, rather than getting them passed from the Nagios server.

That is, in the remote nrpe.cfg, there are lines like: So I guess in the XI config, I could change "Monitor the service with this command" to "check_nrpe!check_load" and leave out the parameters.
Or I could change all the command definitions plus set "dont_blame_nrpe=1" in the remote nrpe.cfg

Sigh....

Anyway, advice is appreciated. Thanks....Lyle

For the performance graphs, XI should automatically generate them for any service that is producing valid performance data. There is also an option in the Core Config Manager to tell Nagios to get performance data. However, if the check NRPE is coming back with an error or no results there won't be any graphs since it doesn't have any performance data to work with.

For check_nrpe you can either have the arguments defined on the Nagios server in the check command, or on the client machine.in the nrpe cfg definition. It can work either way. NRPE is functionally the same in XI as it is in Nagios core, you just might have to define check commands in XI to match your existing configuration.