golang-observability-opentelemetry

Distributed tracing in Go uses OpenTelemetry to create spans that track requests across services. Initialize a tracer provider, configure an exporter, and instrument your HTTP or gRPC handlers to automatically propagate context and create spans for each operation, giving you end-to-end visibility into request flow.

Prometheus metrics collection in Go involves defining metric types—counters, gauges, and histograms—then exposing them via an HTTP endpoint. This enables scraping by monitoring systems to track performance, latency, and error rates across your service in real time.

Structured logging with slog creates context-aware logs in JSON format, making them machine-parseable and easier to correlate with traces and metrics. Configure slog handlers to include request IDs and contextual fields so logs are searchable and traceable across your system.

Yes, OpenTelemetry integrates with Prometheus for metrics and slog for structured logging, allowing you to unify tracing, metrics, and logs in a single observability pipeline. This approach correlates all three signal types across your microservices without replacing existing tools.

Yes, the skill includes implementing liveness and readiness probes for health checks and coordinating graceful shutdown to drain connections and close resources cleanly. This ensures your service handles Kubernetes restarts and rolling deployments safely.

Production gRPC and HTTP services require auto-instrumentation middleware for tracing, metric collection on request latency and errors, structured logging tied to request context, and health endpoints. This combination provides visibility into service behavior and enables rapid incident response.