To ensure that the Secondary host can always detect communication loss regardless of update activity, the Primary host periodically sends a heartbeat message to the Secondary. If the Secondary misses a fixed number of heartbeat messages, it detects a communication loss and tries to reconnect. The reconnecting process triggers the heartbeat protocol. Likewise, if the Primary is unable to send a heartbeat message or if its heartbeat messages go unacknowledged, the Primary also detects a communication loss and enters its recovery procedure. Heartbeat messages use the UDP protocol for communication.

On successful completion of the heartbeat protocol, update activity resumes automatically unless some interim administrative command or error prevents it.