OpenTelemetry for Busy People
summary
OpenTelemetry is explained as a practical observability stack that unifies traces, metrics and logs using language SDKs, a central collector with exporters, and a common wire protocol. A live demo runs a Docker Compose reference app (Python and Java frontends, Spring Boot services) writing data to Postgres, exporting spans to Jaeger and metrics to Prometheus with only a few SDK annotations; Jaeger queries show real traces. The talk notes production concerns around collector sizing and storage growth and mentions common staging approaches like Fluentd/Fluent Bit, Grafana/Loki, or commercial vendors.
AI-written summary · about
chapters (14)
- 0:00OpenTelemetry basics
- 1:49Three pillars of observability
- 2:54Origin and components
- 6:42Signals and specs
- 9:09Practical example
- 13:48Resources and certs
- 15:18Live metrics demo
- 19:39Python and Java code
- 22:20Prometheus integration
- 24:29Q&A
- 27:36Production challenges
- 30:57Implementation strategies
- 33:30Storage and tools
- 34:12Closing
related talks
22mOpenTelemetry – An Observability Framework for Cloud-Native Software
50mGuide to Observability with OpenTelemetry
50mOpen Telemetry: The Rise of Open Standards in Observability
32mOpenTelemetry feat. logs, metrics and traces
38mThe State of OpenTelemetry
43mWhat is this OpenTelemetry thing?
41mModern Application Debugging: An Introduction To OpenTelemetry
37mObservability For You And Me With Opentelemetry