Incidents and maintenance
Tell visitors what is going on. Incidents are for things that go wrong, maintenance is for work you plan ahead. Both show on the status page with a timeline of updates and the services they affect.
Incidents
Post one under Incidents → New incident: a title, the affected services, a severity and a first update.
| Severity | Meaning |
|---|---|
| Minor | Something small, most people won't notice |
| Major | A clear problem for many people |
| Critical | A service is unusable. Its days turn red and count as downtime |
Each update has a status, which moves the incident along:
| Status | Meaning |
|---|---|
| Investigating | We are looking into it |
| Identified | We know the cause |
| Monitoring | A fix is out, we are watching |
| Resolved | Everything is back to normal |
Real start and end times
An incident keeps the time it really started and ended, not the time you typed it in:
- Fill in Ended at to add an incident that is already over.
- Date an update back to when it happened. Updates dated back send no notifications.
- Move the end afterwards when it turns out to be wrong.
Maintenance
Plan maintenance under Maintenance → Schedule maintenance with a start, an end and the affected services. Visitors see it coming on the status page.
| Status | Meaning |
|---|---|
| Scheduled | Planned, not started yet |
| In progress | Work is happening now |
| Completed | Work is done, ends it now |
| Cancelled | Not going ahead |
Cancelled maintenance can be brought back.
Notifications
Posting, updating and resolving incidents, and maintenance that is planned, starts, ends or is cancelled, can go to Discord, email and the bell. Link previews of an incident or maintenance page show its live status when you share it.
Who can post, edit or delete incidents and maintenance is set with roles.