Distributed tracing — Jaeger, AWS X-Ray and trace propagation

Distributed tracing answers the question that neither metrics nor logs can answer alone: 'why is this specific request slow?' Metrics reveal that P99 latency spiked; logs show error messages from various services; but only a distributed trace shows the complete causal chain—which service called which other service, how long each hop took, where the request stalled, and which database query caused the 2-second delay. In a microservices architecture where a single user request crosses five services and makes twelve downstream calls, distributed tracing is the only mechanism that assembles the complete picture.

Two distributed tracing tools are dominant in different deployment contexts: Jaeger, the CNCF-graduated open-source tracing backend widely used with OTel in self-managed Kubernetes environments, and AWS X-Ray, the managed tracing service for AWS workloads that integrates with Lambda, API Gateway, ECS, and other AWS services. Both implement the same core concepts—spans, traces, and context propagation—and both support the W3C Trace Context standard, enabling traces to cross the boundary between Jaeger-instrumented and X-Ray-instrumented services.