Using PMM to identify and troubleshoot problematic MySQL queries
summary
Uses RED: rate errors duration with PMM Query Analytics 2.9 to find which MySQL statements drive load, showing 62 distinct queries, one averaging 120 ms while another causes a p99 800ms spike around 6:40 and worker2 contributes most of the latency. PMM stores query samples in ClickHouse, recommends enabling the extended slow log and setting slow_log=0 to capture IO read ops and lock times so you can sort by duration, load, or disk reads.
AI-written summary · about
chapters (16)
- 0:00Intro
- 0:43Michael Coburn, Principal Architect
- 1:51MySQL Queries
- 5:56MySQL Query Performance
- 6:22USE Method
- 7:38RED Method
- 10:02Query Rate
- 10:51Total Error Rate
- 11:45Total Query Latency
- 13:33Rows Examined per Query
- 15:06Queries Taking the most Time (slowest queries)
- 16:57Queries Causing the greatest Load
- 17:55Queries Waiting on Disk
- 20:14Queries Reading from Disk
- 20:37Query Detail
- 21:29Schema Information
related talks
29mAdvanced Optimization and Troubleshooting MySQL Using PMM2
01h 02mHow-to MySQL Performance Diagnostic Using Percona PMM
58mBetter MySQL With Percona PMM
38mAdvanced PMM Usage
30mWhy we recommend PMM to our clients
50mHow To Improve PostgreSQL Monitoring and Observability With PMM - Part 2
28mThe "411" on PMM
55mLet's Talk About MySQL Query Tuning