Issue 045
Hey folks, welcome to another installment of Monitoring Weekly! Did you write something about monitoring recently? Maybe got an idea rolling around in your head? Send it on over and let the community learn from you. :D
Monitoring News, Articles, and Blog posts
Making 30x performance improvements on Yelp’s MySQLStreamer
A lot of posts we read about here are about someone’s monitoring stack or some cool new tool/approach, but this one is a little different: how Yelp actually used monitoring, logging, and code profiling tools to improve performance of a particular application.
Do you version your Grafana dashboard configuration? Did you even know you could? It’s a little bit of work but I like the idea.
Our journey implementing Sensu to monitor Kubernetes in production
Pushing Sensu client data to Prometheus and Kubernetes, resulting in a really neat monitoring stack. Pretty cool stuff.
Full disclosure: My company, Aster Labs, is a Sensu Partner. I received no consideration, financial or otherwise, for including this post.
Graphite at Scale at Criteo: BigGraphite Bye bye whisper
A talk from FOSDEM 2018 (which just ended, so great job on getting this posted so quickly FOSDEM folk!) on Graphite storage when you’ve got 80+ million metrics, a read rate of 20k metrics/second, and a write rate of 800k metrics/second. They set out to replace Whisper but keep Carbon and Graphite-Web. The audio isn’t the best, but this is a good talk and worth watching.
This is an interesting thing. First off, it’s an app published by The National Association of Realtors, who aren’t exactly known for their software engineering prowess. Second, it leverages some of our favorite tools (Grafana!) to present data collected from…your house.
It’s not “monitoring at Dropbox” (though that would be an awesome post too–you hear that Dropbox?), but rather “how I’m monitoring my Dropbox.” The article is written by someone at Circonus (a monitoring company), so it makes heavy use of Circonus, but the approach is easily translatable. I think I might set this up myself, even.
SRE@Xero: Managing Incidents Part II
I featured Part I in Monitoring Weekly a couple months back, and now Part II is here. This part goes into much more detail on their Hubot-based chatops incident management system–how it works, the functionality it provides, and more.
Woooo! It’s been a few years since Monitorama Berlin (2013!) but now it’s confirmed for Amsterdam. Should be an awesome event and a great opportunity for you Europe folks to get to a Monitorama event easier.
See you next week!
– Mike (@mike_julian) Monitoring Weekly Editor