Trace

Trace — in Monitoring

Definition: A trace represents the journey of one request through a system, especially when it passes through multiple services.
It lets you see which operations happened, how long they took, and where a problem occurred.

Examples:

1. User → API → Payment Service → Bank → Database.
2. A page takes 5 seconds, and the trace shows the database query took 4.5 seconds.
3. An order fails in the Payment Service, and the trace identifies exactly where it failed.

trace / tracing in observability.

A trace follows one request through the whole system.

Simple example

A user sends:

POST /orders

The backend gives that request a unique trace ID:

trace_id = abc123

Then everything caused by that request carries the same trace ID:

Request
  │
  ├── Controller
  │      trace_id = abc123
  │
  ├── Service
  │      trace_id = abc123
  │
  ├── Database query
  │      trace_id = abc123
  │
  ├── Call payment service
  │      trace_id = abc123
  │
  └── Send notification
         trace_id = abc123

So later you can say:

“Show me everything that happened during request abc123.”

Trace vs log

A log is an individual event:

ERROR Payment service failed

A trace connects many events together:

Request abc123
    ↓
Auth
    ↓
Create order
    ↓
Database
    ↓
Payment API
    ↓
Error

What are callbacks?

If your request triggers asynchronous work, you generally want that work to remain associated with the original operation where appropriate.

For example:

HTTP Request
    │ trace_id=abc123
    ▼
Create Order
    │
    ├── DB query
    │
    └── Queue message
          │
          ▼
       Worker
          │ trace/context propagated
          ▼
       Send Email

This lets you investigate the entire chain, rather than seeing unrelated logs from different requests.

One important terminology detail

A trace usually contains multiple spans:

Trace: abc123
│
├── Span: HTTP request
├── Span: database query
├── Span: payment API call
└── Span: email operation

So:

• Trace = the whole journey of one request/operation.

• Span = one operation within that journey.

• Trace ID = identifier connecting the spans.

• Tracing = the practice of recording and following this journey.

This is commonly implemented with OpenTelemetry and is one of the three major observability signals:

Logs + Metrics + Traces.

---

My-Journey-In-Codeless Back-End