Monitoring
31 practice items · needed by 9 roles
Easy 11Medium 10Hard 10
Practice questions
- Crash spike during a releasehard
- Alerting on high CPUeasy
- How Prometheus collects metricseasy
- What Grafana is foreasy
- Alarm silent when the app diesmedium
- Requests per second in PromQLmedium
- Broken p95 latency queryhard
- Fixing alert fatigue with SLOshard
- The four golden signalseasy
- How Prometheus collects metricseasy
- The USE methodeasy
- The RED methodeasy
- Percentiles vs average latencymedium
- Combining percentiles from many servershard
- Where to measure the SLIhard
- Seeing what is running noweasy
- Measuring replica lagmedium
- What data drift meanseasy
- Inputs stable but accuracy fallshard
- Users start asking new thingsmedium
- What data drift iseasy
- What p99 latency meanseasy
- Labels arrive weeks latermedium
- What to log for each LLM callmedium
- Fleet-wide p99 from many podsmedium
- Online features stopped updatingmedium
- Drift alerts firing all the timehard
- Model affects its own training datahard
- GPU shows 100% but is underusedhard
- Detecting stale source datamedium
- Tests that run in productionhard