Redbrick's nagios setup
These should all be physical machines, and are all found in
on halfpint (One config file per host). Service IPs should not be setup as a seperate host but as a
service on the correct host.
nrpe is installed on all of the machines, and is checking load and hard disk space at a minimum. On
morpheus, it's also checking ircd, and on severus, it's checking that the two mysqls are in sync.
All of the service checks are in
halfpint. For things which have service ips, define a command to check the service ip (in
/usr/local/etc/nagios/objects/Redbrick/commands/commands.cfg) and add it as a service under the
host it runs on.
Adding a new machine
If you're adding a new machine, then add it to
/etc/nagios/objects/Redbrick/machines/, and make it
a member of redbrick-login, redbrick-routers or redbrick-services as appropriate. This will add ping,
disk, and load checking (you'll need to set up disk and load checking with nrpe on the host side
first of course). It won't set up ping checks on the external interface however, you'll have to do
that seperately, as well as any services that the machine runs.
After updating anything
After making your changes/additions you'll want to have nagios verify the config files are
nagios -v /etc/nagios/nagios.cfg
Assuming nothing went horribly wrong you can restart nagios,the rc script is of course in
/usr/local/etc/rc.d instead of
/etc/rc.d because of superior FreeBSD file layout or whatever.