The offline alert

The offline alert is the simplest and most important alert. It tells you when your server has stopped responding for long enough that it is likely down, not just momentarily unreachable.

What triggers it

The alert fires when no heartbeat is received for 5 minutes. With a 3-minute polling interval, this means two consecutive failed polls before the alert sends. The delay is intentional: brief network interruptions and server restarts typically resolve within two poll cycles. Requiring two consecutive failures avoids alerting on planned restarts and transient connectivity issues.

A missed poll vs going offline

A single missed poll is not an alert. It is recorded in your uptime history, affecting your uptime percentage, but does not trigger a notification. Two consecutive missed polls flip the server to "offline" status and send the alert.

If your server restarts frequently, you may want to suppress the offline alert during known maintenance windows. This is available from the Alerts settings in your dashboard: set a maintenance window and alerts will be paused for that period.

What the alert contains

The email includes the context from the last successful heartbeat before the server went silent:

• Last known TPS and MSPT
• Last known player count and player list
• Last known heap usage
• Time since last heartbeat

The last-known TPS and memory values are useful for diagnosing whether the server crashed under load, with low TPS or high memory just before going silent, or stopped unexpectedly while performing normally, which suggests an external cause like a host issue or a scheduled task killing the process.