Alerting
Route different servers and checks to different people, with timing rules that match how you actually want to be interrupted.
Five channels
Configure one once and reuse it — channels belong to the project, not to a single check.
On every plan with nothing to configure.
Webhook
A structured payload to any endpoint you control.
Slack
Into the channel your team already watches.
Discord
For teams whose operations channel lives there.
Telegram
A message or channel post, straight to a phone.
Group the channels, then attach the group
A notification policy is a named set of channels with its own timing rules. Attach it to a server or check in one step, and change where everything routes by editing one thing.
Delay
Wait before the first alert, so a single bad check that resolves on the next run never reaches a person.
Repeat
Keep reminding on your interval while it is still broken, so a real outage does not go quiet.
Recovery
Say when it is fixed, through the same channel that said it broke.
Why it stays quiet
- An alert waits for the number of confirming regions you set, so one bad route is not an outage
- Repeated failures on the same subject roll up into one incident rather than one message per check
- Pending updates and reboot-required are recorded and never alerted on, because they are true on most hosts most of the time
Start monitoring in about a minute.
Twenty checks and one server, free and with no card. Upgrade only when you outgrow it.