In the first part to this article I showed how I setup Nagios on a host and configured it to monitor itself.  Not  particularly useful in and of itself,  but it was a good starting point.  In this next article I’ll show you how I’m using the Nagios host to monitor any number of remote machines.

When I originally wrote these notes I was setting up this system on mostly Fedora 11 and 12 machines.  Since then I’ve done the same setup on both Fedoara 13 and 14 machines and a couple of Windows 2003 servers with very little modifications, I’ll try to point out any differences as we go.

The first step is to install NRPE (Nagios Remote Plug-Ins Executor) on the remote machine you want to monitor, along with the neccessary plugins.   I chose to install nagios-plugins-all just because I use most of them on at least one server or another, so it simplified the setup proccess accross all our servers.  You may want to have a look at what plugins are available and choose only the ones you’ll need, the last time I checked there was quite a few.

You can view the available list using YUM like so:

Once you’ve decided what ones you’ll need go ahead and install them, here’s what I did:

The next step is to setup the nrpe daemon to run under xinetd. I should mention here that you don’t have to use xinetd to run nagios, it can be run unter inietd or as a stand-alone daemon if my memory serves me correct, this is the way I decided to run it however.  I believe with Fedora 11 and 12 xinetd isn’t installed by default so you may need to do so:

Now you can create/edit the nrpe config file under xinetd:

Add the following to the file:

Note: make sure you add the ip address of you host nagios machine next to ‘only_from = 127.0.0.1′.

The next step is to add nrpe somewhere in the services config:

Now restart xinetd:

Now check that everything has been configured correctly by using the following command:

Which should display something like this:

If it displays nothing then the service isn’t setup, go back and make sure you didn’t miss any steps, otherwise continue on.

If you have a firewall setup make sure you open port 5666. If you’re using iptables you can do something like this:

If you want to make sure that the ip is added successfully you can list your current iptables setup:

If all’s well you should see the nrpe service listed like so:

At this point you might want to have a look at the nrpe config file for nagios located at /etc/nagios/nrpe.cfg. You shouldn’t need to change any of the defaults for the purpose of this post as we’re only going to monitor ping, ssh and http for this machine. However if you want to monitor some of the internal processes and statistics like swap size and disk space you’ll have to configure what Nagios refers to as ‘command definitions’. I may make another post with regards to this later but in the mean time if you want to add this functionality have a look at the bottom of the file under “COMMAND DEFINITIONS”.

In this next step you’ll need to open a terminal on your nagios host machine and add the new remote machine to the nagios configs.

Start by creating the config file for the new remote machine we just configured.

Insert the following definitions in the the config:

We’ve done two things in this config file, 1 we’ve defined the remote server for nagios inside the define host definition, and 2 we’ve define 3 services that we’re going to monitor on the remote server, ping, ssh and http. As I mentioned above there are a lot more services and statistics that you can monitor, but these should be a good starting point.

Now we need to add a reference to this new config file in Nagios’ main config.

Now you can run a sanity check on the nagios host config:

If all’s well then you should get some output with this at the end:

If everything is fine you can start the nrpe daemon on the remote machine:

Then restart nagios on the nagios host machine so it can see the new configurations

Gotchas:

If you get a response like:

Make sure that you have dont_blame_nrpe=1 in your /etc/nagios/nrpe.cfg file on the remote server.

If you get a response like:

Make sure you have the corresponding command in your /etc/nagios/nrpe.cfg file on the machine running nrpe (the remote machine). For the error above the entry to fix it would look something like this:

If you’ve made it through and there’s no errors then you should be able to open up your Nagios admin page and see the new remote server in your list of hosts. Happy monitoring!

————————————————————

References:

Nagios Quick Start
Nagios Core Documentation