Why architecture matters here
Oncall fails on alert noise, poor runbooks, missing handoff, and no burnout care. Architecture matters because sustainability + reliability are one system.
The architecture: every piece explained
The top strip is coverage. Rotation — primary + secondary. Alerts — actionable only. Escalation clear. Runbook for common ops.
The middle row is learning + care. Handoff context. Postmortem blameless. Burnout metrics — interrupts + hours. Comp / time off compensate.
The lower rows are hygiene. Alert quality review. Team size minimum 6 for humane rotation. Ops — training + shadow + feedback.
End-to-end flow
End-to-end: oncall week begins. Handoff meeting reviews open issues. Alerts fire; runbook guides mitigation. High-severity escalates. Postmortem after major incident. Metrics review reveals 40% of alerts unactionable; team reduces noise. Comp reflects duty.