What alerts you'll get
Every message we send, what triggers it, and how to get fewer of them.
The messages
Down. A monitor's consecutive failures reached its threshold. Names the monitor and the error — "connection refused", "keyword not found in response", "certificate expires in 9 days".
Recovered. The first successful check after a down. Includes how long it was down.
Slow response. Response time exceeded your latency threshold for the number of consecutive checks you set. The monitor is still up: this is information, not an outage.
Content changed. A content change monitor saw the page change. One per change, and never a down.
Cannot fetch. A content monitor has failed to fetch its page for about a day. Without it, a watched page that started 404ing would simply go quiet.
Who receives them
Whichever channels the monitor uses, or every team member by email if you have configured none.
Getting fewer alerts
In the order worth trying:
Raise the failure threshold on anything that flaps. Most alert noise is one dropped packet being treated as an outage.
Check the right thing. A keyword monitor on a page that includes a third-party widget will alert when the widget's host has a bad day. Watch something only your own systems control.
Use maintenance windows for planned work. Deploys that take a site down for forty seconds generate an incident and two alerts every time, and people learn to ignore them — which is far more dangerous than the deploy.
Separate the urgent from the interesting. Latency alerts and content changes rarely need to reach a phone at night. Give them a different channel from down alerts.
Prefer Ping to nothing, Port to Ping. A ping monitor on a host that drops ICMP will alert forever about nothing at all.
Getting more alerts
If something broke and you were not told:
- Was the monitor paused, or in a maintenance window?
- Does the monitor have a channel attached, and did the test pass?
- Did the failure last long enough to reach the threshold? A 30-second outage on a 5-minute interval may not produce a single failed check.
- Was it a webhook channel whose receiver was also down? There are no retries.
Last updated 4 October 2026
Still stuck?
If this did not answer your question, tell us and we will fix the page as well as answer you.