I want to setup a monitoring solution whose overall purpose is to be able to fix server problems. What tool do you guys find best for doing this?
I've been leaning towards Nagios, because it has the notifications set up and has a few plug ins, but it still doesnt feel like it is all there, especially when it comes to performance metrics.
Error monitoring: Rollbar (http://rollbar.com) - yes, we use it on itself
Performance monitoring: Rollbar, Pingdom, and some custom instrumentation