Where it is raised
POST /api/v1/incidents/{id}/updates
What triggers it
An update that sets the incident to resolved looks at the live state of the incident's component and refuses a component still measured as down. Sending force true skips that comparison. The message names the check, the region when there is one, and whether the last reading was down or degraded.
Example response
POST /api/v1/incidents/{id}/updates
HTTP 409
{
"error": "Our last probe still reports <check name> from <region> down. Marking this incident resolved tells every subscriber and every status page visitor that the outage is over. Send force: true to resolve anyway.",
"code": "RU-5003"
}Common causes
- Resolving an incident before the underlying monitor has actually recovered.
- Automation resolving on a timer rather than on recovery.
How to fix it
- Wait for the monitor to recover: the incident can then resolve normally.
- If you are certain the probe is wrong, pass `force: true` to override; the override is deliberate, never the default.