In order to achieve a high level of availability for the CAS role, there are two requirements: deploy more than one CAS and use a load balancing solution. Both these requirements have the purpose of decreasing the downtime for users by allowing failover of connections in case of an issue with one or more CASs. This failover is often automatic, depending on the load balancing solution implemented.
Load balancing is the method by which incoming client connections are distributed over a variety of endpoints (in our case, CASs) to ensure the load across all endpoints is similar. In Exchange, from an availability perspective, its main purpose is to provide failover redundancy if one or more CASs fail, ensuring that users continue to be able to access their mailboxes, thus reducing the impact of a CAS failure.
Furthermore, both these advantages are often achieved while presenting a single hostname for clients.
In Exchange 2010, the CAS role handled and processed client...