skill-observability-tracing

Implement OpenTelemetry distributed tracing by initializing instrumentation libraries early in your Node.js and Python applications to capture spans and export them to an OTLP endpoint.

Context propagation passes trace context across service boundaries, ensuring distributed traces are connected. OpenTelemetry provides propagation utilities to maintain this correlation end-to-end.

Export OpenTelemetry traces to Jaeger or Tempo by configuring an OTLP exporter in your tracing setup. This sends span data directly from your instrumented services to the backend.

Yes, you can apply manual span usage alongside semantic conventions. This ensures consistent tracing metadata across distributed services for accurate latency diagnosis and error tracking.

Distributed tracing sampling strategies filter the spans exported to your backend. OpenTelemetry allows configuring sampling to balance visibility against performance overhead in microservice architectures.

Distributed tracing correlates traces and spans across multiple services, providing end-to-end visibility. This allows you to pinpoint exactly where latency occurs within a microservice architecture.