Last Friday going into the weekend I ran across a snapshot on one of my VMware hosts almost 160 days old, OUCH. The right tool to keep that from happening is definitely Nagios. NagiosExchange didn’t really have a solution for my problem that I could find. Somebody has written a snapshot age tool in PowerShell but I’m not interested in having plugins run on hosts that aren’t my main Nagios server. I was given a fun project to work on.

The vSphere Command Line Interface (formerly the PERL toolkit if I’m not mistaken) was of little help. It didn’t really give me any interface into snapshot data at all. I decided the simplest solution would be to work right on the BusyBox console. I started Friday around noon and working on it here and there over a couple days came up with a usable product yesterday morning:

[jrdalrymple@nagios ~]$ /usr/local/nagios/libexec/check_snapshot.py
No password specified
usage: check_snapshot.py -H hostname [-U username] <-P password | -f PasswordFile>

[jrdalrymple@nagios ~]$ sudo /usr/local/nagios/libexec/check_snapshot.py -H 172.16.100.11 -U nagioschk -f /home/nagios/.check_esxi_hw.pw -w 10 -c 20

3 VMs are CRITICAL
Guest example1.domain.local has snapshot 24 days old!
Guest example2.domain.local has snapshot 28 days old!
Guest example3.domain.local has snapshot 26 days old!

The results between my command line run and the Nagios GUI aren’t the same because I gave the Nagios check different thresholds.

I’ll probably put it up on Nagios Exhange at some point. For now I’ll just feel accomplished.