MySQL High Availability Deep Dive: Pacemaker, DRBD, MySQL Replication, and more!
In this popular tutorial first presented at this year's MySQL Conference and Expo in Santa Clara, Florian explores the Pacemaker cluster stack, fully integrated with MySQL. He covers DRBD based and MySQL replication based cluster solutions, and combined configurations — all within one high availability framework.
In a 2-node high availability cluster, we walk you through setting up DRBD storage replication, Pacemaker high availability, and MySQL as a highly available, failure monitored, automatically recovering cluster resource. This is the simplest, quickest and easiest way to set up MySQL high availability in Pacemaker clusters, and can be accomplished by an experienced sysadmin in an hour or less. We then cover multi-node MySQL replication under Pacemaker management. With this, you can ensure quick and seamless failover with any MySQL storage engine, making use of MySQL replication for scale-out or redundancy. Finally, we dive into solutions combining high-availability with scale-out and disaster recovery. We show how to deploy mixed DRBD/MySQL replication configurations, with both transaction-safe high availability, and load-distributing scaleout.
Attendees should bring laptops with VirtualBox installed to follow the practical sessions in the tutorial. You will be provided with pre-installed virtual machine images running 32-bit Linux. Required memory per virtual machine is approximately 512MB. For the full deployment count on running 4 virtual machines in parallel (total memory: 2GB). The virtual machines will run on any reasonably recent hardware.