Incident.
also: incidents · severity level · severity levels · SEV1 · SEV 1 · SEV2 · SEV 2An unplanned disruption or degradation of service that needs an immediate response, graded by severity from SEV 1 (critical) to SEV 4.
the two-minute refresher
- Not every alert is an incident: CPU at 75% is monitoring; error rate 5% against a 1% SLO is.
- SEV 1: complete outage for all users, all hands. SEV 4: minor, normal ticket queue.
- Lifecycle: detection → triage → response → mitigation → resolution → post-incident.
- Mitigate first (rollback, failover); root cause comes later.