Architecting an open observability stack
summary
Argues for an open-source observability stack built around OpenTelemetry for vendor-neutral tracing/metrics/logs, with concrete backends: Jaeger/Zipkin for traces, Prometheus + Grafana for metrics and dashboards, and Grafana Loki or Logstash for logs. Recommends using auto-instrumentation SDKS to start, instrumenting application functions, exporting traces/metrics, and rewriting PromQL to surface latency, error rate and request rate. Describes a prototype that generates those queries and plugs into Grafana.
AI-written summary · about
chapters (7)
- 0:00Introduction to observability
- 1:30Requirements for observability
- 4:19Reliability and SLOs
- 7:09Vendor dependency in observability
- 9:17OpenTelemetry standard
- 10:41Auto-instrumentation benefits
- 11:44Designing an open stack
related talks
42mOpen Source for Better Observability
01h 02mDesigning an open-source observability stack for Kubernetes
45mOpen Source for Better Observability
14mObserving the Future
35mOpenTelemetry for Busy People
28mOpen Source for Better Observability
37mBuilding Modern Observability Architecture With OpenSearch, OS Ingest...
34mThe Grafana Stack: Building a Highly Available, Scalable Observability Platform