You need to ensure that App1 fails over automatically to another server if a single node in Cluster1 fails