The Foundry switches feature best-of-breed intelligence to maximize server utilization, protect capital investment, and allow for infrastructure growth on demand. The Server Iron switches act as traffic cops in front of the DNS servers, and ensure high service uptime and superior scalability. They efficiently distribute client requests to the "best" server in the pool. The switches consider server availability, load and response time in server selection to provide the best end-user response time and service availability. Using sophisticated "health checks" to DNS servers, the ServerIron switches identify unavailable resources in real time and switch users to other available servers. They also help scale server farms on demand by adjusting server capacity without impacting application availability. When more capacity is needed, a new server can simply be added to the pool without impacting the end user. The ServerIron switches feature advanced modes of high availability to extend service resilience across the failures of a ServerIron switch. In the high availability mode, two ServerIron switches are deployed to back each other up during failures.