PMM Beyond Monitoring Observe, Understand, Act
summary
PMM expands from metrics to a closed loop that observes with OpenTelemetry support, understands via an AI assistant that produces root‑cause analysis, and acts through a new management framework that runs predefined, tested jobs. First apps are MySQL backups and diagnostic data collection; planned features include MongoDB bootstrap, backup/restore, and rolling upgrades. Alerts become fewer and more actionable; LLMs provide analysis but the framework enforces deterministic jobs for execution. Releases: management framework next, OpenTelemetry after, assistant later.
AI-written summary · about
chapters (10)
- 0:00PMM introduction and evolution
- 2:15Limits of traditional monitoring
- 3:40Observe: OpenTelemetry metrics
- 6:30Understand: AI root cause analysis
- 7:36Rethinking alerts
- 8:40Act: AI agents framework
- 10:25Trust in automated actions
- 11:50Management tools and apps
- 13:14Database management options
- 14:15PMM roadmap and features
related talks
28mThe "411" on PMM
38mAdvanced PMM Usage
22mGetting Started with Percona Monitoring and Management
17mOpen Source Database Operations in the Age of AI
58mBetter MySQL With Percona PMM
04mMongoDB Monitoring & Alerting with PMM
21mEasy way of application monitoring
50mHow To Improve PostgreSQL Monitoring and Observability With PMM - Part 2