Best Practices for MySQL Performance
summary
A monitoring and troubleshooting toolkit is presented that combines Prometheus and Grafana for the frontend, Consul for configuration, and a custom query analytics component; the client exporter collects OS and database metrics from MySQL/MariaDB/Percona/Amazon RDS and MongoDB. The talk gives a concrete case where a query spent 70% of its time in lock wait, and describes distribution via Docker/AMI, open-source licensing, low performance impact, and an option to operate without exporting data; release 1.4 follows a year of development.
AI-written summary · about
chapters (13)
- 0:00Introduction
- 2:11Performance causes
- 5:06Who suffers
- 8:45Problem-solving methods
- 11:39Database performance issues
- 13:49Monitoring databases
- 16:45Monitor overview
- 19:44Metrics analysis
- 22:49Query problem detection
- 24:24Query optimization
- 27:06Extending monitoring
- 28:41Custom metrics
- 30:51Q&A
