Sign in Start free

← Docs  /  Watchtower uptime

Incidents & downtime alerts

Updated 2026-09-05

How an outage becomes an incident, who gets emailed, reminder cadence, and flap control.

Browse documentation

What it does

An incident opens after two consecutive failed checks. WPClientHub waits, then re-checks before mailing — so a site that recovered in the meantime never triggers a false alarm. Reminders during a long outage back off rather than repeating constantly: 30 minutes, 2 hours, then 6 hours after the first alert, then every 12 hours until the site recovers. A flapping site (opening and closing incidents every couple of minutes) is capped at one "down" email per site per window, paired with exactly one matching recovery email; every suppressed outage is still recorded, and the next email that does go out says how many were held back.

How to

  1. Configure who receives downtime alerts under Notifications (personal preference) or notification routes (an external email, Slack channel or webhook).
  2. Open an incident from the Uptime tab's incident list to see its timeline and which region(s) reported it down.

Limits & defaults

SettingDefault
Incident opens after2 consecutive failed checks
Confirm-before-alert delay~3 minutes
Minimum outage to alert on~2 minutes
Reminder cadence30 min, 2 hr, 6 hr after first alert, then every 12 hr
Flap controlAt most 1 down-alert per site per 30-minute window

Troubleshooting

No one got the downtime email
Check that the alert kind is on for at least one person in Notifications, or routed externally — see Emails not arriving.