Trace Types

To understand what Tracing Service can show, start with the kinds of operations that distributed tracing typically captures.

• Server spans: These represent inbound requests handled by a service, such as an HTTP API receiving a client request.
• Client spans: These represent outbound calls from a service to another service, such as HTTP requests or RPC calls to downstream dependencies.
• Internal spans: These capture in-process work inside a service, such as validation, business logic, or background processing steps.
• Producer spans: These represent operations that publish messages, events, or jobs to asynchronous systems such as queues, topics, or streams.
• Consumer spans: These represent operations that receive and process messages from asynchronous systems.
• Error traces: Traces can include span status, events, and attributes that indicate failures, retries, or degraded operations.
• Cross-service traces: The most valuable tracing workflows combine many span types into one distributed trace, allowing you to follow a request across APIs, services, databases, and asynchronous components.