Operations, Maintenance, and Troubleshooting
What this domain tests
Operations questions test whether you can keep a system running, detect degradation early, diagnose faults methodically, and change systems safely.
Concepts to master
Understand monitoring vs observability - metrics, logs, and traces - and the four golden signals: latency, traffic, errors, saturation. Know SLI, SLO, and SLA and how error budgets work. Understand alerting on symptoms rather than causes, runbooks, on-call escalation, and incident severity levels. For maintenance, distinguish corrective, adaptive, perfective, and preventive maintenance. For troubleshooting, know a layered method: reproduce, isolate by layer, form a hypothesis, test one variable at a time, confirm the fix, then write it up.
Traps in questions
The classic trap is jumping to a fix before isolating cause - options that restart services first are usually wrong unless the question is explicitly about restoring service. Distinguish mitigation (restore service) from resolution (remove root cause); both are correct answers to different questions. Another trap is alerting on a cause (CPU high) rather than a symptom (users see errors).
Worked rule of thumb
Ask whether the question wants service restored or cause eliminated. If users are down, mitigation wins even if it is inelegant. If the question says "prevent recurrence", the answer is root-cause analysis and a permanent change.
Change safety
Maintenance changes need rollback plans, staged rollout, and verification against the golden signals. A change with no rollback path is a risk decision, not a routine operation, and should be treated as such in any question about approval.