There are two key functions within CQ5 dispatcher that should be highlighted:

• Caching: The CONTENT on the website is cached as much as feasible in this role so that it can be used as a static website. It will lessen the layout engine's need to run repeatedly. The Dispatcher caches static items in a Cache Directory. The cached documents are created in the web server's root directory. For caching, Dispatcher EMPLOYS the following techniques:
  • Content Updates invalidates pages with updated content and replaces them with fresh content.
  • Without physically removing any files, auto-invalidation automatically invalidates out-of-date content portions.
• Load balancing: Load balancing is the process of dividing the website's computational load across many INSTANCES of AEM. This is used to improve the website's performance. This is how Dispatcher performs load balancing:
  • Performance Statistics — Dispatcher keeps track of how quickly each instance of cq responds to a given url. Using these metrics, dispatcher determines which instance of cq will fetch the fastest response for any request and forwards the request to that instance.
  • Sticky Connections — Once a user session has been established, all incoming requests from that user should be serviced by the same cq instance, as other cq instances will not be able to identify the user session and provide personalised pages for him. The dispatcher ensures that all user session requests are served from the same cq instance.
  • Increased fail-safe coverage: If an instance does not respond, the Dispatcher will immediately relay requests to one of the other instances (s)
  • Increased processing power: In reality, this means that the Dispatcher distributes document requests over MULTIPLE CQ instances. You get faster response times since each instance has fewer documents to process.