A proxy cache server intercepts HTTP requests from clients, and if it finds the requested object in its cache, it returns the object to the user. If the object is not found, the cache goes to the object's home server - the originating server on behalf of the user, gets the object, possibly deposits it in its cache, and finally returns the object to the user.
Proxy caches are usually deployed at the edges of a network (i.e., at company or institutional gateway/firewall hosts) so that they can serve a large number of internal users. The use of proxy caches typically results in widearea bandwidth savings, improved response time, and increased availability of static web-based data and objects.
Stand-alone proxy configuration has one disadvantage that is the cache represents a single point of failure in the network. When the cache is unavailable, the network also appears unavailable to users. Also, this approach requires that all user web browsers be manually configured to use the appropriate proxy cache. Subsequently, if the server is unavailable (due to a long term outage or other administrative reason), all of the users must re-configure their browsers in order to use a different cache. One final issue related to the 1 approach has to do with scalability. As demand rises, one cache must continue to handle all requests.