distributed-tracing-specialist

Distributed tracing provides end-to-end visibility into request flows across services. Initialize OpenTelemetry with a tracer, create spans for operations, propagate trace context via W3C headers over HTTP and Kafka, and correlate traces with logs and metrics in Jaeger for operational insight.

W3C Trace Context propagation embeds trace identifiers into HTTP headers, enabling spans across service boundaries. Extract context from incoming requests, inject it into outgoing calls, and use middleware to automate this process for all HTTP framework operations.

Yes. OpenTelemetry supports trace context propagation over Kafka by extracting and injecting context into message headers, allowing traces to flow through async queue operations and maintaining correlation across producer and consumer spans.

Embed trace IDs from OpenTelemetry spans into log output and metric attributes, creating a unified observability signal. Jaeger links these signals together, enabling drill-down from traces to logs and metrics for complete operational context.

Middleware integration captures HTTP requests and responses as spans automatically without manual instrumentation. Configure tracing middleware in your HTTP framework to record span data, record errors, and propagate context without code changes per endpoint.

Automatic middleware captures HTTP layer spans; manual spans instrument business logic, database queries, or custom operations. Combine both: let middleware handle HTTP, then create explicit spans for domain-specific work to maintain trace granularity.