Status pages
Incidents & maintenance
How incidents are opened, updated, and resolved, and how to schedule maintenance windows.
Automatic incidents
Section titled “Automatic incidents”When a monitor’s failure is confirmed, StatusCheck opens an incident, sets the affected status-page components to the matching state, and notifies channels and subscribers. When the recovery is confirmed the incident is resolved the same way. You never have to remember to update the page for something the monitors can see.
Automatic incidents start with a factual first update covering which component, since when and the failure reason, written in plain language. Edit it, or add updates, at any time.
Manual incidents
Section titled “Manual incidents”Open an incident from Incidents → New incident for problems your monitors can’t observe: a degraded feature, a data delay, a third-party outage. Choose the affected components, their state, and write the first update.
Updates and the timeline
Section titled “Updates and the timeline”An incident is a timeline of updates, each with a status:
| Status | Use it when |
|---|---|
| Investigating | You know something is wrong and are looking. |
| Identified | You know the cause. |
| Monitoring | A fix is in place; you’re watching. |
| Resolved | It’s over. |
Every update is posted to the status page and sent to subscribers. Keep them short and specific: what is affected, what you know, when the next update will come. A template library is available in the editor for the updates you find yourself writing repeatedly.
Post-incident
Section titled “Post-incident”After resolution you can add a postmortem to the incident, which is a longer write-up of cause and follow-ups. It appears on the incident’s page in the status-page history. Publishing it is optional.
Maintenance windows
Section titled “Maintenance windows”Schedule a maintenance window from Maintenance → Schedule:
- Pick the affected components and monitors.
- Set the start time and duration.
- Write the notice that subscribers receive.
Subscribers are notified when the window is scheduled, when it begins, and when it ends. During the window the status page shows Under maintenance for the affected components, and monitors covered by it do not alert. Checks still run and are recorded, so the uptime history stays honest, but nobody is paged for downtime you planned.
Windows can repeat weekly or monthly for recurring maintenance.