Monitoring MySQL with Prometheus
The Prometheus timeseries monitoring system is a perfect match for collecting detailed metrics from MySQL. Prometheus can provide both performance analysis and health alerting. A single Prometheus server can monitor millions of metrics for thousands of schemas over hundreds of servers with minimal effort. In my talk I will give a short overview of Prometheus and timeseries monitoring techniques. Then I will give a deeper dive into the features and possibilities of monitoring MySQL with Prometheus and the mysqld_exporter. About me: I am a Site Reliability Engineer/Systems Engineer at SoundCloud. I have over 15 years of systems experience and have been involved in high performance computing/supercomputing for over 10 years.
Site Reliability Engineer/Systems Engineer, SoundCloud
I've been a sysadmin/systems engineer since founding a tiny web hosting ISP in 1996. Since 2002 I have been focusing on high performance computing. Starting with Linux compute clusters at the University of Minnesota, then as a Borg/GFS SRE at Google from 2005 through 2013. In 2013 I moved to Berlin, Germany to join SoundCloud. I had never used MySQL for anything but hobby projects. I have since become the company MySQL tech lead, applying SRE techniques to MySQL service management I am the current maintainer of the Prometheus mysqld_exporter. The mysqld_exporter is a FOSS tool that allows Prometheus to collect, graph, and alert on MySQL internal status metrics.