A probe that quietly goes down at 3am is no better than no probe at all. Alerts are the half of monitoring that reaches you: every time a probe changes state, Uptimeprobe records it and sends it wherever you’re watching.
There are two sides to it. The Activity tab on your Alerts page is the running history of what happened. Destinations is where those changes get delivered.
What creates an alert
A probe moving between the three health states creates one entry:
| Change | What it usually means |
|---|---|
| Up to Down | The thing you’re watching stopped working. |
| Up to Degraded | It still works, but it’s flaky or slow. |
| Degraded to Down | It got worse. |
| Down to Up | It recovered. |
| Degraded to Up | It’s healthy again. |
Each entry records which probe changed, the status it moved from and to, and when the change was detected.
What doesn’t create an alert
Three of the six statuses carry no signal about the service you’re watching, so moving into them is deliberately silent:
- Scheduling. A probe you’ve just added, before its first check. Adding a probe never alerts anyone.
- Paused. You turned the probe off, so you already know.
- No data. Checks have stopped arriving, so Uptimeprobe can’t say whether the service is healthy. Reporting that as an outage would cry wolf.
When checks resume, the probe picks up from its last known health state. A quiet gap that starts and ends with the service up produces no entries at all, rather than a false Down followed by a false recovery.
Because No data is silent, a probe that stops reporting entirely won’t notify you today. Check the Activity tab if a probe has gone unusually quiet.
Which changes get sent
Each destination chooses which statuses it wants. New destinations start on Down and Up: you hear about the outage and the recovery, and nothing in between.
Adding Degraded is useful for a service where slow is as bad as broken, but expect more traffic — an intermittent endpoint can move in and out of Degraded repeatedly while it misbehaves. Every change still lands in the Activity tab either way, so filtering a status only changes what gets delivered, never what gets recorded.
Where alerts go
Right now Uptimeprobe delivers to webhooks: an HTTPS endpoint you control, which lets you route alerts into whatever you already use for on-call. See Destinations to set one up, and Webhook payloads for what arrives and how to verify it.
Next steps
Set up your first destination in Destinations, or read Probe statuses for what each state means and when a probe moves between them.