추적 스팬과 ID 이해

Welcome to Distributed Tracing! Ever wonder exactly what happens when a user clicks a button, and that request travels through many services?

Traditional monitoring struggles with this. Distributed tracing is your solution to understand the full journey of a request across complex systems.

Think of a trace as the complete story of a single request or operation as it flows through all the services in your application.

• It starts when a request enters your system.
• It ends when the final response is sent.
• It captures every step in between, no matter how many services are involved.

A trace is made up of smaller units called spans. Each span represents a single, distinct operation or unit of work within the trace.

For example, fetching data from a database, calling another microservice, or executing a specific function could each be a span.

Every span holds important information about the operation it represents:

• Operation Name: What happened (e.g., authenticateUser).
• Start/End Time: When the operation began and finished.
• Duration: How long it took.
• Attributes (Tags): Key-value pairs providing context (e.g., http.method: GET, user.id: 123).

The Trace ID is a unique identifier assigned to the entire request journey. All spans that belong to the same request share the exact same Trace ID.

This ID is crucial for linking all related operations together, allowing you to reconstruct the full story of a request.

While the Trace ID identifies the whole story, the Span ID uniquely identifies a specific operation (span) within that story.

Each span gets its own unique Span ID. This helps distinguish individual steps in the trace, even if they happen concurrently.

Spans aren't just a flat list; they form a hierarchy. When one operation triggers another, they become parent-child spans.

A child span will typically reference its parent's Span ID. This creates a tree-like structure, showing causality and nested operations.

These IDs and relationships allow observability tools to visualize a trace as a waterfall diagram or a tree. You can see:

• The total time taken for the request.
• Which services were called.
• How long each step took.
• Any errors that occurred at a specific step.

Let's look at a conceptual Java example. Imagine a request being handled, which then calls an authentication service and a database service. Notice how the Trace ID is constant, but Span IDs change, and parent-child relationships are indicated.

Great job! You've learned the fundamental building blocks of distributed tracing:

• A Trace is the complete journey of a request.
• Spans are individual operations within a trace.
• The Trace ID uniquely identifies the entire request.
• The Span ID uniquely identifies an individual operation.
• Spans form parent-child relationships to show causality.

These concepts are key to understanding how distributed tracing provides deep visibility into your complex systems. Next, we'll explore how tracing actually works!