Quiz: Tracing

← Back to quiz index

2 questions

L1 (1 questions)

1. When should you use metrics, logs, or traces for debugging?

Show answer

Metrics: detect and alert on symptoms (high error rate, latency spike). Good for dashboards, trends, and capacity planning. Logs: understand what happened in a specific component (error messages, stack traces). Good for debugging known services. Traces: understand how a request flowed across services and where it slowed down. Good for debugging cross-service latency and dependency issues. Start with metrics to detect, traces to locate, logs to diagnose.

L2 (1 questions)

1. A request is slow but each service reports low latency. How do you find the bottleneck?

Show answer

Distributed tracing. Check spans for gaps between services (network/queue delay), fan-out (parallel calls serialized), or missing instrumentation hiding a slow hop.