Monitoring Redis/Resque

Real quickie post tonight! I realized that the last two projects I’ve forked on Github are both for monitoring Redis… and both are adding password support. Using Redis’ built-in auth seems to be fairly uncommon within the community.

Tonights patch was to check_resque, a clever Nagios plugin for watching the length of various queues and alerting on given thresholds. It’s the perfect way to provide some critical notification if workers are getting behind. The plugin is just querying Redis for the size of the lists in question using Perl’s Redis library which added password support a few months back. Grab check_resque and throw it in your Nagios checks dir, ignore the makefile business. Add it to your commands config ala $USER1$/check_resque.pl $ARG1$ and invoke it via check_resque!-H <host> –password <redis password> –queue <resque queue> -c <critical threshold> and you’re done!. See -h for more options.  Respect to its author, it’s well written.

The previous fork I’d done was to redis-munin, a fantastic set of metrics for Redis installs using the Munin suite. It’s pure bash and will give you a multitude of graphs about Redis activity and performance: clients, commands, hits/misses, blocked clients, changes and key eviction. The docs have example configuration, but it’s mostly drop in and go. I think it’s pretty fantastic, so grab it and get some insight in your Redis install. All kudos to its author too.

That’s all, busy day, have a good night!