First Steps with Observability: OpenTelemetry and You
summary
Concrete steps for getting observability: deploy the OpenTelemetry collector to centralize headers and stitch distributed traces, set memory limits and batch sizes to prevent metric explosion, and route metrics to Prometheus/Grafana or vendors like Honeycomb or TelemetryHub. Distributed tracing relies on consistent trace headers and baggage for cross-service context; many frameworks offer auto-instrumentation (Java, Node, Rails, .NET). One surprising datapoint: moving some services back to a monolith cut monitoring costs by about 90%.
AI-written summary · about
chapters (15)
- 0:00Observability and OpenTelemetry intro
- 2:29Microservices observability challenges
- 6:39Three pillars of observability
- 11:16Tracing as observability pillar
- 14:28Distributed tracing consistency
- 16:57OpenTelemetry Collector explained
- 19:04High cardinality storage issues
- 23:37Demo: Node.js instrumentation
- 28:50OpenTelemetry libraries stability
- 29:51OpenTelemetry baggage testing
- 31:14Storage and scaling
- 34:01Greenfield observability
- 36:53OpenTelemetry open standard
- 38:39Auto-instrumentation adoption
- 40:26Brownfield telemetry import
related talks
57mYour Distributed System is Talking to You: OpenTelemetry
36mObservability For You and Me with OpenTelemetry
22mShift Into an Observability Mindset with OpenTelemetry
26mYou Can’t Fix What You Can’t See: Tales from the Observability Frontier
37mObservability For You And Me With Opentelemetry
35mOpenTelemetry for Busy People
22mOpenTelemetry – An Observability Framework for Cloud-Native Software
50mOpen Telemetry: The Rise of Open Standards in Observability