Road to Percona Live 2022 - Ananias Tsalouchidis
summary
Builds a centralized alerting pipeline that plugs Alertmanager + PMM together to deduplicate group and route alerts using metrics already collected in PMM and to reduce noise by detecting sustained anomalies rather than single spikes. Demonstrates a concrete rule that fires when connections remain elevated over a 30-minute pattern window, shows routing to PagerDuty OpsGenie email and claims fewer false positives, with the integrated system actually configured and exercised end to end.
AI-written summary · about
related talks
25mUsing PMM Alerting in Production
28mThe "411" on PMM
22mGetting Started with Percona Monitoring and Management
37mSmart Data-driven Alerts (SDdA) and How They Make DevOps Life Easier
15mPMM Beyond Monitoring Observe, Understand, Act
29mAdvanced Optimization and Troubleshooting MySQL Using PMM2
51mHow to improve PostgreSQL Monitoring and ...
30mWhy we recommend PMM to our clients