intermediate
Runtime operations
Operate deployed services with health checks, probes, logs, metrics, alerts, and basic incident response discipline.
Runtime operations interviews cover keeping services healthy in production: probes, structured logs, metrics tied to SLOs, actionable alerts, and disciplined incident response.
Subtopics: health checks, readiness/liveness probes, logs, metrics, alerts, incident basics.
On interviews: design `/health/ready` for an API with a database; choose one SLI and alert; walk through first 15 minutes of a checkout outage.
Common pitfalls: identical liveness and readiness; logging PII; high-cardinality metrics; paging on every error log.
The trade-off is observability coverage versus cost, noise, and on-call sustainability.
Checklist:
- Separate readiness from liveness.
- Correlate logs and metrics.
- Alert on symptoms and SLO burn.
- Run blameless incidents with follow-ups.