Case Study: Deploying a Multi-Region, Highly Available MySQL Architecture
summary
Designs a three-region Azure MySQL deployment with a single-master (chosen for lowest latency) and local read replicas in each region, traffic routed by Azure Traffic Manager and site VPNs. Uses ProxySQL in a middle layer for read/write splitting, weighted read failover, TTL query cache and comment-based rules to force read-after-write to the master. Uses MHA as topology manager with cloud API fencing to promote a new master on region failure. Works for a read-heavy app but lacks proxy-level HA and requires cross-region write hops.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 1:13Requirements and environment
- 3:23Data replication strategies
- 5:59Proxy layer design
- 8:24Technology selection
- 10:57Topology management
- 13:21Final architecture
- 15:19Failure scenarios
- 17:17Limitations and considerations
- 19:14Q&A session
related talks
24mComparing the High Availability of MySQL with Modern Distributed SQL Databases
08mMySQL as a Layered Service: How to use Proxy SQL to Control Traffic and Scale Out
35mMulti-Site, Multi-Master Done Right
52mBest Practices for MySQL High Availability in 2022
47mHigh Availability in GCE
32mBest Practice in Design and Implementing MySQL Geographic Distributed, High Availability Solutions
12mDave Stokes Talks Open - Episode 2
51mAchieving Resiliency With A Multi-Cloud Or Multi-Region Approach