Nagios virtualization

As virtualization seems to be a trendy thing to do, I went ahead and virtualized our nagios (while reinstalling the whole thing …).

Now as I went into work today and started my email client, I received 4 nagios warnings about a LOAD service reaching critical state. Looked at the nagios box itself, opened up the VM console, looked into the syslog. Nothing.

Yet over 3/4 of the services were flapping, some ping checks were critical (for whatever reason). So I opened the nagios webinterface again, and noticed it dropping the connection over and over again (had to reauthentificate me again and again).

So I opened up Putty, which established the connection without a single problem, but dropped me like a stone after a short amount of time. I restarted the session and got a security warning from Putty (due to different than the saved sshd public key). That raised my suspicion. So I took a look at the hostname, and lookie there.

Somehow my old nagios box (which is a physical box), got turned online again, thus having the same IP address as my virtualized one. So the virtualized nagios wasn’t really dropping my connection, but I was being directed to the old nagios.

Walked over into the data center, turned of the old box (well, I kept the power button pressed for a short time), and away went my troubles.