Scale MySQL beyond limits with ProxySQL
summary
Zapier swaps vertical RDS scaling for a centralized proxy layer and chooses ProxySQL over HAProxy Nginx MaxScale because it is transaction-aware, supports connection pooling and query routing. Benchmarks with sysbench show roughly 10x reads and 5x writes under load while maintaining far fewer DB connections; production sees primary connections dropped ~1,500 → ~300 and about $1,500/month saved. Deployment uses a highly available ProxySQL cluster, monitoring of pool/connection metrics, and workarounds for long transactions and logging.
AI-written summary · about
chapters (11)
- 0:00Start
- 2:10Shortcomings
- 2:56Duct Taping
- 4:23Available solutions: Proxies
- 4:39Available solutions: MariaDB MaxScale
- 5:43Our Choice?
- 10:36How much can it multiplex?
- 11:06Zero downtime configuration update
- 12:16How does it work with (Django) ORM?
- 13:54ProxySQL & SAVEPOINTS
- 15:59Deployment strategies
related talks
50m#ProxySQL Use Case Scenarios
08mMySQL as a Layered Service: How to use Proxy SQL to Control Traffic and Scale Out
48mMySQL Load Balancers - MaxScale, ProxySQL, HAProxy, MySQL Router & nginx - a close up look
55mProxySQL Use Case Scenarios
06mRoad to Percona Live 2022 - Oleksandr Kuzminskyi - ProxySQL Talk
54mProxySQL at Scale
27mIntroduction to Configuring and Using MariaDB MaxScale: Database Proxy Made Simple
25mThe future of databases: Combining operational and analytical data at massive scale