Measure What Matters
Talkat Observability Day North America 2024 · on Tuesday, 2024-11-12· at 14:05· Salt Palace | Level 2 | 255 E·
summary
Explains how Honeycomb plus OpenTelemetry auto instrumentation plus selective custom spans turns noisy CPU and throughput pages into actionable signals: the team turned off throughput alert for an upstream-caused page, adds span attributes (geo, device, app version) and a stress level attribute on Refinery to page only on sharp cluster stress spikes, and sets SLOs like checkout 2s at 99.5% so alerts map to real user impact and many nuisance pages become async tickets during business hours.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 1:05Reducing alert noise
- 5:59Actionable alerts with telemetry
- 7:02Custom instrumentation
- 10:56Span attributes for context
- 12:40Service level objectives
- 16:55Aligning SLOs with business
- 18:20Evolving SLOs
- 21:30Recap and next steps
- 22:53Q&A
related talks
50mObservability and the Development Process
45mObservability is for User Happiness
42mFounder/CTO Perspectives: The Future of Distributed Tracing
40mDecoding Synthetic Monitoring: A Journey from E2E UI Tests to Service-Level Probes
31mObservability for Developers: How to Get There
19mOops my app performance sucks... what if observability could be a game changer?
53m.NET Rocks Live
36mMetrics - What's your code actually doing?