Glossary
heartbeat
An inverted check — your service pings the monitor on a schedule, and the monitor alerts when the pings stop.
A heartbeat (also called a heartbeat monitor or dead man’s switch) is an inverted check: instead of the monitoring service probing your application, your application pings the monitoring service on a defined schedule. The monitor watches for the pings to stop. If a ping is more than a configured tolerance late, the monitor fires an alert.
Heartbeats are the right primitive for monitoring cron jobs, scheduled tasks, queue workers, and other “should run every N minutes” processes. They detect the absence of work as cleanly as a normal check detects the absence of a response.