Last week I spent a bit of time scratching an itch I’ve had with doing maintenance. We use Nagios for our monitoring (without any addons, plugins or frontends) and scheduling downtime is a pain. Not only that, but deleting downtime is impossible short of screenscraping or (shock) actually clicking in the user interface. When I’m done doing maintenance I want to delete all the downtime (regardless of if the window will naturally come to an end soon or not) because I want notifications to start sending immediately if there is a real problem.

Naturally I did some diligence beforehand as this is hardly a topic that hasn’t come up before, and sure enough some guy has implemented something similar but in Python. Nevertheless I wanted to do some coding so I present to you Sinagios.

It’s a pretty simple Sinatra app (and thus obviously Ruby), in fact it only does three things:

• Lists currently scheduled downtime
• Schedules new downtime
• Deletes downtime

I also got to dive into a bit of RSpec and Rack-test since I’ve mostly gotten away with Test::Unit before and non-web-facing systems so it was an interesting change. Fairly soon I’ll be uploading a spec file for making RPMs and adding some Puppet recipes to deploy it so that it’s a more complete set of stuff.

Hopefully someone will find it useful!