Heartbeat monitoring
A heartbeat monitor expects your system to check in regularly. If it goes quiet you get a notification, and another one when it comes back.
In the dashboard open your project → Heartbeats → New heartbeat monitor: name, interval (e.g. 1 hour), grace period, priority and sound. You get a private ping URL:
# crontab: after the backup finishes0 3 * * * /usr/local/bin/backup.sh && curl -fsS https://api.pushbell.app/v1/heartbeat/<token> > /dev/nullGET, POST or HEAD work; no API key is needed (the URL acts as the key, keep it private).
States
Section titled “States”- Waiting for a ping: no ping yet; no alerts.
- Up: pings arrive on time.
- Late: last ping + interval (+ a small network margin) has passed.
- Down: the grace period has passed too → notification.
- Paused: paused by hand (e.g. during maintenance); no alerts.
The ping time is when the request reached us. Even if our side has an outage, pings wait in the queue and are processed with their real times, so an outage does not cause false alarms.
Device inactivity
Section titled “Device inactivity”On My devices, pick an “Inactivity alert” for a device: if it does not open or sync the app for that long (say, the shop tablet was switched off), your other devices are notified, and again when it is back.