Measured status, not assumed uptime
Check times and results are stored. Gaps without measurements are not shown as successful checks; uptime is based only on checks that actually ran.
Downtime and recovery
By default, three consecutive failures open an incident and two consecutive successful checks resolve it. A single temporary network error does not immediately become an incident.
Settings per domain
Set the check path, expected HTTP code, timeout and frequency for each domain. Checks and incident notifications can be paused during scheduled maintenance.