Open Telemetry for API Monitoring
summary
OpenTelemetry instruments APIs with traces spans and logs, uses context propagation (W3C trace context) and a collector that receives processes and exports telemetry to backends like DataDog or Grafana, enabling SLI/SLOs and dashboards built on the four golden signals. Practical guidance shows where to inject spans — controllers databases HTTP clients — and deployment options for exporters. Case study: AWS outage: nine hours traced to internal DNS and an automatic scaling job, OpenTelemetry collector as agent or sidecar, instrument controllers DB and HTTP clients.
AI-written summary · about
chapters (8)
- 0:00Introduction to OpenTelemetry for API monitoring
- 1:09Why API outages happen
- 4:08Need for observability
- 7:41OpenTelemetry fundamentals
- 15:35Collector and data flow
- 17:26Dashboards and golden signals
- 19:39AWS outage case study
- 21:28Q&A
related talks
46mThe State of OpenTelemetry
50mGuide to Observability with OpenTelemetry
22mOpenTelemetry – An Observability Framework for Cloud-Native Software
24mObservability and instrumentation via OpenTelemetry
30mOpenTelemetry 101
21mOpen Telemetry
15mExtending OpenTelemetry to Observability
59mOpenTelemetry will save your day and night