What is distributed tracing?

Short answer

Distributed tracing tracks the flow of requests through microservices. It captures timing and metadata for each hop, enabling root cause analysis of latency and failures.

How it works

A trace ID is generated at the entry point and propagated across service boundaries. Each service creates spans representing operations. Spans are collected and assembled into a trace.

Example

from opentelemetry import trace

tracer = trace.get_tracer(__name__)

with tracer.start_as_current_span("process_order"):
    process_payment()
    update_inventory()

Why it matters

In microservices, a single request may touch dozens of services. Tracing reveals bottlenecks and errors that logs and metrics cannot capture in isolation.

Tips

• Use OpenTelemetry for vendor-neutral instrumentation.
• Sample traces to reduce overhead.
• Correlate traces with logs and metrics.

Common issues

• Incomplete instrumentation leaves gaps.
• Trace propagation requires consistent headers.
• Storage costs grow with trace volume.