dapr-observability-setup

End-to-end observability for DAPR requires configuring OpenTelemetry tracing, Prometheus metrics, and structured JSON logging. Call configure_tracing() in your service, set up an OTLP exporter to route traces to backends like Jaeger or Azure Monitor, start a Prometheus metrics server, and enable JSON log formatting for centralized analysis across distributed services.

OpenTelemetry tracing captures the full request path across services with span processors and auto-instrumentation for FastAPI and HTTP clients, showing latency and dependencies. Metrics collection using Prometheus exposes quantitative data like request counts and response times. Together they provide request-level visibility and system-level performance monitoring.

Yes. OpenTelemetry integrates with DAPR via OTLP export to Jaeger, Azure Monitor, Prometheus, and other backends. DAPR configuration blocks control tracing sampling and endpoint connections, while Python instrumentation applies tracer providers and span processors to your services.

Yes, structured logging requires JSON formatting and instrumentation in your service code. The Skill provides configuration templates and Python logging setup that integrate with your existing DAPR services without breaking changes, improving log traceability across microservices.

You need a DAPR project, a metrics backend (Jaeger, Prometheus, or Azure Monitor), Python environment with FastAPI or HTTP service, and network connectivity to your observability endpoints. DAPR configuration blocks and Python instrumentation then route traces and metrics to these backends.