For example, if a cluster was started with three systems (System-A, System-B and System-C) in that order, then System-A will be the Coordinator since it was the first to join the cluster. If System-A fails for any reason, then System-B will become the new Coordinator. If System-A is restarted and successfully rejoins the cluster, then the order of the systems (for becoming a Coordinator) would then be System-B, System-C and then System-A. If System-B were to crash, then System-C would become the new Coordinator.

It is possible for the clustering communications channel between the systems to break due to network interruptions or other reasons.

Managed File Transfer

has built-in heartbeats and other safety checks to overcome small network interruptions. However, if the communication becomes unresponsive for 60 seconds, then the Participant systems will believe that the Coordinator system has failed. When that happens, each Participant will start checking the database to see if the Coordinator system is still running.

If the Coordinator system is still running after 30 seconds (if the communications channel fails) then the Participant systems will automatically shut down. This is required in order to prevent issues that can occur when multiple

Managed File Transfer

systems are running on the same database and not communicating with each other. In this rare scenario, even though the Participant systems were shut down, the Coordinator system will still be up and running.