Monitoring our systems is a problem that is not yet solved. That is why initiatives like Monitoring Sucks 1 exist.
Riemann 2 is a a very powerful monitoring system that helps us monitor our distributed systems. With riemann we can use a powerful stream processing language to analyze our metrics and send emails, track metrics, combine statistics, etc.
But switching to riemann is not easy if you have a statsd/graphite monitoring architecture in place. You would need to send all your metrics to riemann instead of statsd and this takes some time to change.
In this talk I will give an introduction to riemann and show how you can set it up without changing your existent statsd/graphite architecture. You can continue to send metrics to statsd and they will be processed by riemann so your #monituringsucks less.
My name is Simão Mata. I was born in Portugal and moved to Berlin three years ago. For the last year I have been working as backend/devops engineer at wimdu.com, where we have some very interesting problems to solve. I write about software at https://simao.io/blog.