observability-and-instrumentation

Implement distributed tracing with OpenTelemetry by integrating the Node SDK to visualize request flows across service boundaries, enabling precise diagnosis of which service failed during an incident without manual log parsing.

RED metrics track Request, Error, and Duration rates for services, while USE metrics track Utilization, Saturation, and Errors for resources. Both are necessary to establish symptom-based alerting and avoid high-cardinality data points.

Add correlation IDs to structured JSON logging by propagating trace context across service boundaries, ensuring logs remain queryable and traceable across distributed components during production incident investigations.

Auto-instrumentation for Node.js applies to services, endpoints, and background jobs requiring reliable diagnostic evidence. It automatically captures telemetry without needing manual code changes for standard library integrations.

Avoid high-cardinality labels in Prometheus metrics to prevent database bloat and query performance degradation. This skill guides tracking RED/USE metrics while maintaining cardinality limits for symptom-based alerting.

Review existing service logging by analyzing structured event tracking implementations against RED/USE metrics and distributed tracing requirements, suggesting improvements to ensure reliable diagnostic evidence for production operations.