Observability for Developers
summary
Autometrics is an open source microframework that instruments functions and emits Prometheus/OpenTelemetry metrics so developers get three metrics per function (invocation rate, error count, duration histogram) and automated P95/P99 calculations. SLOs are attached in code so SLOs defined in code generate recording rules via Sloth and feed any Prometheus/OTel backend; a simple UI groups metrics by function/module and the demos show it working with multiple vendors and SLO-based alerts.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 1:28Mental models
- 4:18Pets to cattle shift
- 6:23Left shift in SDLC
- 7:48Mental models vs monitoring
- 10:12Autometrics framework
- 13:18Developer-first metrics
- 14:01Q&A
- 17:39Closing remarks
related talks
53mOpenTelemetry Metrics: observability’s holy grail
29mObservability; a gentle introduction
20mObservability and Open Telemetry
48mAPI Observability by Design: Building Systems That Tell Their Own Stories
50mGuide to Observability with OpenTelemetry
38mObservability 101 with Spring and Micrometer
56mFull-stack observability and application monitoring with
22mOpenTelemetry – An Observability Framework for Cloud-Native Software