Distributed Tracing

This document describes how an SDK should propagate information between different services to connect all telemetry (errors, profiles, replays, transaction, check-ins, ...) from those services into one trace.

For an overview see Distributed Tracing in the product docs.

Sentry uses three containers to hold trace information sentry-trace, baggage and optionally traceparent.

With these containers you can propagate a trace to a down-stream service by either:

• adding sentry-trace and baggage HTTP headers (when making outgoing HTTP requests),
• adding sentry-trace and baggage as meta data (when putting tasks into a queue, details are specific to the queue you want to support), or
• setting environment variables (when calling another process). In this case the env variables should be called SENTRY_TRACE and SENTRY_BAGGAGE.

The SDK running in the receiving service needs to make sure to pick up incoming trace information by

• reading sentry-trace and baggage headers for each incoming HTTP request,
• reading sentry-trace and baggage meta data when retrieving an item from a queue, or
• reading the environment variables SENTRY_TRACE and SENTRY_BAGGAGE on startup.

This trace information should be stored in the "propagation context" of the current scope. This makes sure that all telemetry that is emitted from the receiving service to Sentry will include the correct trace information.

W3C Trace Context Header

To interoperate with OpenTelemetry (OTel) services, SDKs can optionally send the W3C Trace Context traceparent HTTP header on outgoing requests in addition to Sentry’s own sentry-trace and baggage. This enables traces originating from a Sentry SDK to be continued by components that only use OTel/W3C propagation.

The traceparent header must only be sent when propagateTraceparent is true and the request matches tracePropagationTargets (same condition as for sentry-trace/baggage).

The full spec is available in the W3C Trace Context document.