patternjavascriptModerate
Status page design: communicating incidents to users and stakeholders
Viewed 0 times
status pageincident communicationStatuspageCachetuptime historyRCA post-mortemsupport volumeseparate domain
Problem
During an incident, users and stakeholders flood support channels with 'is the system down?' questions, overwhelming the engineering team that is trying to resolve the incident. There is no authoritative source of truth for system status.
Solution
Maintain a public status page that is updated within minutes of incident detection. The status page should be hosted independently of your main infrastructure (it must work when your service is down).
Tools:
Status page components to include:
Communication principles:
Tools:
- Atlassian Statuspage — most widely used, integrates with PagerDuty
- Instatus — affordable alternative with good UX
- Cachet — self-hosted open-source
Status page components to include:
- Each major product area (API, Dashboard, Webhooks, Data Pipeline)
- Status for each region if multi-region
- Current incident banner with timeline and next update time
- 90-day uptime history per component
Communication principles:
- Update within 5 minutes of detection with 'Investigating'
- Never say 'all systems operational' during an active incident
- Give next update time: 'Next update in 30 minutes'
- Post RCA (root cause analysis) within 48-72 hours of resolution
- Subscribe link for email/SMS/webhook notifications
Why
A public status page reduces support volume by 40-60% during incidents. It signals transparency and builds customer trust. The page should be updated by a dedicated communications role, not the engineer fixing the issue.
Gotchas
- Host the status page on a separate domain and infrastructure — statuspage.io, not status.yourdomain.com on the same servers
- Automate status updates from your alerting system to reduce manual work during incidents
- Never mark an incident as 'Resolved' prematurely — wait until you have confirmed full recovery
- Include the status page URL in your API error responses: {'error': 'Service unavailable', 'status': 'https://status.example.com'}
Context
Setting up incident communication infrastructure for a production service with external users
Revisions (0)
No revisions yet.