We Monitor Everything… Except What Matters
summary
Describes operating AWS systems where abundant metrics, dashboards and alerts still failed to explain latency, intermittent errors and silent degradations because teams monitored the wrong things. Argues that many alerts track symptoms not root causes and metrics are disconnected from real user impact, and recommends criteria for choosing what signals to collect so monitoring helps troubleshooting under pressure rather than creating noise.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 1:05Real learning stories
- 3:12Initial monitoring focus
- 8:52Useless monitoring examples
- 12:25Business-focused shift
- 16:57Team discovery process
- 19:00Costly lessons learned
- 24:41Importance of fixability
- 25:21Q&A
related talks
39mThe Ticking Timebomb of Observability Expectations
38mThe Ticking Time Bomb of Observability Expectations
18mHow not to practice observability
52mHolistic Monitoring
05mMonitoring Decoded, Why What and How? in 5 Ignite minutes
23mMastering Monitoring & Alerting: Key to Seamless Performance
30mThe Four Questions (Every Monitoring Engineer gets asked)
16mUnderstanding Alert Fatigue