Incident Communication

Lifecycle states

Incidents move through:

1. investigating
2. identified
3. monitoring
4. resolved

Severity levels

• minor: limited scope, low customer impact
• major: broad impact or significant degradation
• critical: severe service interruption

Update cadence

• Initial update as soon as practical after detection.
• Follow-up updates at regular intervals while impact continues.
• Resolution update when service metrics recover.
• Postmortem publication for impactful incidents.

Channel model

• /status as canonical public channel
• RSS feed for status updates
• Optional email subscriptions (when enabled)

Postmortem expectations

A postmortem should include:

• incident summary and impact window
• root cause
• mitigation timeline
• prevention actions and owner follow-up

Example timeline format

• 09:12 UTC - investigating
• 09:37 UTC - identified
• 09:54 UTC - monitoring
• 10:03 UTC - resolved