The mode in which Application Request Routing is deployed on two ARR servers that process requests and that together act as a failover server. This configuration can achieve high availability, but is still limited by the maximum capacity of one ARR server.