Incidents
An incident is the record of an outage — when a monitor went down, how long it stayed down, and when it came back. TunnelHQ opens and closes them automatically from your check results.
On this page
The lifecycle
When a monitor transitions to down (or degraded), TunnelHQ opens an incident and stamps its start time. While the outage continues, that same incident stays open — it doesn't spawn a new one every cycle. When the monitor recovers, the incident is closed with an end time and moves to recovered.
- Down — the monitor is failing. For multi-region monitors this follows the aggregate, so the incident reflects all regions, not one.
- Degraded — partially reachable: some regions pass, some fail.
- Recovered — the monitor came back; the incident is closed.
Start time is the current outage
The Incidents page
Incidents in the sidebar lists everything for the current project, grouped into ongoing and recovered. The summary cards at the top show active outages, anything needing attention, acknowledged incidents, and the recovered count. Filter by time range, status, or search by server, and export to CSV when you need a record.
Acknowledging
Acknowledge an incident to signal that someone's on it. An acknowledged incident records who took it and an optional note, so an outage that's being handled is visibly distinct from one nobody has looked at yet.
Getting notified
Incidents are the trigger for alerts. Wire up a webhook to receive a clean monitor.down the moment one opens and monitor.up when it recovers, or send notifications through the channels under Settings → Notifications.