Incident Reports
Root cause first. Blame last.
Your Mission
A DevOps engineer has produced the following timeline after a production outage: "14:32 — Deployment of v2.4.1 started (automated pipeline) 14:38 — Deployment completed successfully, health checks passed 14:41 — First error alerts: database connection timeouts on checkout service 14:43 — PagerDuty escalation, on-call engineer (Priya) paged 14:47 — Priya identifies spike in database connections, suspects connection pool exhaustion 14:52 — Decision to roll back to v2.4.0 14:55 — Rollback initiated 15:01 — Rollback complete, services recovering 15:08 — All services nominal Total downtime: 27 minutes Affected: checkout service, order confirmation emails Root cause (confirmed): v2.4.1 introduced a missing connection pool size configuration that defaulted to 5 instead of 50, exhausting the pool under normal load within minutes of deployment." Write a prompt that instructs the AI to produce a formal incident report suitable for internal distribution to engineering leadership and the operations team.