Updated 2026-09-29
Each check has its own alert policy. These options apply to every alert channel, whether email or webhook.
Confirmations and the confirmation window
Confirmations set how many regions must agree before a check is down (see How multi-region confirmation works). The confirmation window is how long Status waits for those regions to answer: from 30 to 300 seconds, 60 by default.
Alert delay
An optional wait after confirmation, from 0 to 60 minutes, before the first notice goes out. If the check recovers inside the delay, nothing is sent.
Recovery notices
On by default. A recovery notice goes only to the channels that received the opening alert.
Re-alerts
Professional and Enterprise. While an incident stays open, Status can repeat the alert every 1 minute to 24 hours until it resolves or is acknowledged.
Escalation
Enterprise. If an incident is still open after 1 minute to 24 hours, Status can notify a second set of up to 20 channels. Escalation stops when the incident is acknowledged.
Acknowledging
Acknowledging an incident stops re-alerts and escalation. See Managing incidents.
Slow response
For HTTP, TCP, ping and DNS checks you can set a response-time threshold (1 to 120,000 ms) and how many runs in a row must exceed it (1 to 20, 3 by default). When that happens Status opens a separate "degraded" incident.
Timeout bursts
On by default: 5 timeouts within 5 minutes sends a burst alert. You can change the count (1 to 100) and the window (1 to 60 minutes), or turn it off. It is sent at most once per window, and not while the check has an open incident.
Flapping
When a check changes state 6 or more times in an hour, you get one "flapping" notice (at most every 30 minutes). The down alert is held for up to 30 minutes, then sent if the check is still down. See Why is my check flapping?.