Requests fail only sometimes across services. What do you use to follow one request?
Why A is the answer
Distributed tracing with OpenTelemetry
Tracing links work across services, so you can see where failures or delays occur.
In detail
- A trace links one request across all services.
- Sample traces, and keep all the errors.
Use case
An illustrative scenario, not a real incident.
One request in fifty fails across three services. Following one failing trace shows which service returned the error.
See it in different ways
flowchart LR A["API"] --> B["Worker"] B --> C["Database"] A --> D["Application Insights"] B --> D C --> D
sequenceDiagram participant A as Service A participant B as Service B participant C as Service C participant M as Application Insights A->>B: Request, with trace context B->>C: Request, with trace context A->>M: Spans B->>M: Spans C->>M: Spans
flowchart LR S0["Find a failed request"] --> S1["Open its trace"] --> S2["Find the failing span"] --> S3["Fix it"]
Key point: Traces follow single requests.
Why the other options are wrong
- B. A single log line in the first service only
- One log line shows one service.
- C. Azure Advisor recommendations
- Advisor gives general recommendations.
- D. CPU metrics for each service
- CPU metrics do not follow a request.