Trace ID
The shared ID that ties every timed step of one request into a single trace, even when the work crosses services.
See it
What it is
A trace ID is the identifier shared by every span that belongs to one trace. In W3C Trace Context it is a 16-byte value written as 32 lowercase hexadecimal characters inside the traceparent header. Each span also has its own span ID, while the trace ID stays constant as the operation crosses functions, processes, and services.
Reach for it when you need to pull one request's entire journey out of a large telemetry store. Put the active trace ID on structured log records too, so a log line can jump to the matching trace and a trace can reveal the nearby logs. OpenTelemetry creates and propagates it automatically when instrumentation is wired correctly.
Gotcha: the chain breaks silently when one HTTP client, queue producer, or worker fails to propagate trace context. Do not generate a fresh trace ID at every service boundary, and do not treat the value as a secret or an authentication token. It is a correlation handle, often copied into logs and URLs.
Ask AI for it
Instrument this service with OpenTelemetry and W3C Trace Context. Accept and validate the incoming traceparent header, keep the same trace ID while creating a new span ID for each operation, and inject traceparent into every outbound HTTP request and queue message. Bind the active trace_id and span_id to every structured log record, expose them in error reports, and add an integration test proving that one trace ID survives an API to worker to database path.