Workspace System Design
Learning pathsA
COURSE CHECKPOINT · 5 QUESTIONS

Quiz: Metrics Monitoring

Apply Metrics Monitoring to concrete decisions. Choose an answer for every question, then review the explanations.

Visual checkpoint

Use this diagram to locate the state and failure boundaries tested by the questions. Answer all five questions before revealing feedback; the result is a study checkpoint rather than a certification.

Instrumented services → Collectors (samples); Collectors → Series ingestion (bounded buffer); Series ingestion → Time-series storage (series + time); Time-series storage → Query / alert engine (window queries); Query / alert engine → Notification state (pending / firing); Series ingestion → Query / alert engine (freshness signal)
Scroll to inspect the diagram, or open it at full size.

Figure — Instrumented services → Collectors (samples); Collectors → Series ingestion (bounded buffer); Series ingestion → Time-series storage (series + time); Time-series storage → Query / alert engine (window queries); Query / alert engine → Notification state (pending / firing); Series ingestion → Query / alert engine (freshness signal)

Review your reasoning

After checking answers, identify the incorrect assumption behind each missed question. Rewrite it as a rule with a concrete exception or failure case, then return to the lesson.

Why is averaging the p99 latency from ten instances not the service-wide p99?

1. Which label is dangerous for metric cardinality?
2. Does missing telemetry mean zero errors?
3. Can ten instance p99 values be averaged to obtain service p99?
4. Why use a pending duration before firing an alert?
5. What happens when a counter restarts at zero?

Answers and your latest score are saved to your account when you check them. Completion records your attempt, not a passing certification.