Core concepts
Understand pings, survival intervals, status URLs, and alert states.
Ping URL
A public URL called by your job after a successful run. It updates the canary's last check-in time and clears a previous failure state.
Max survival interval
The longest acceptable silence window. Use a value slightly longer than the normal schedule plus expected runtime variance.
Status URL
A public read-only URL that returns OK or FAIL. Use it when another monitoring system needs a simple status check.
Choosing a survival interval
| Job schedule | Good starting interval | Why |
|---|---|---|
| Every 5 minutes | 10-15 minutes | Allows one delayed run without hiding a real outage. |
| Hourly | 75-90 minutes | Covers runtime variation and queue delays. |
| Daily | 25-27 hours | Avoids false alarms around clock drift and DST-adjacent scheduling. |