dt-obs-tracing

Diagnose performance bottlenecks by analyzing distributed traces to identify root spans, map service dependencies, and compute latencies across cross-service calls. It extracts actionable insights from standard span attributes to pinpoint critical paths.

Distributed trace analysis requires standard span attributes including trace.id, span.id, duration, span.kind, and dt.service.name. Trace data must contain these fields to support root-span extraction, service lookups, and extrapolation.

You can run DQL-based investigations to query trace data, extract root spans, and analyze service dependencies. This allows you to compute error rates and uncover failure reasons from span data within Dynatrace environments.

This approach supports OneAgent-based traces and Dynatrace environments, applying to end-to-end observability workflows across server-side traces, client spans, database interactions, and message processing.

Map service dependencies by visualizing cross-service calls within distributed traces to identify critical paths. This process highlights bottlenecks and failure points across services to help teams quickly understand application performance.

The best way to analyze failure reasons is to compute latencies and error rates directly from span data. By evaluating service relationships and trace patterns, you can isolate the root causes of failures across distributed services.