Documentation Center

About Content Distribution

Content distribution in Content Delivery lets you use load balancing to horizontally scale your capacity to serve content to your users in a way that keeps the content consistent, up-to-date, and readily available.

The content distribution model in Content Delivery consists of a single master server that delivers packages and publication content to a content distribution group, which consists of one or more content delivery servers. On the master server, system administrators (publication managers, style developers) define distribution profiles containing publication content, skins, and/or application configuration. At set intervals, the master server pushes the contents of these distribution profiles to the content delivery servers, which process the packages to be ready for users to view. This insures that each content delivery server is identical in terms of content, configuration, and appearance, a critical factor in presenting information to your user. Regardless of which content delivery server handles a user's request, the response will be identical.

When a user requests to view a document, the request is received by the load balancer. The load balancer then establishes a session with one content delivery server, and that server will continue to fulfill a user's requests until the session is terminated. The load balancer may choose any of the content delivery servers when a new session is started, and this choice is based on various factors (such as load or network throughput) and the load balancer's configuration.

Content distribution model

The content distribution model allows you to distribute many user requests across multiple content delivery servers. Therefore, each server can handle only a percentage of the total volume of requests, thereby enabling faster response time for your users. For example, if your site has one content server which receives 1,000 hits-per-minute, your site response times would improve if that load were distributed over four content delivery servers, allow each to handle only 250 requests per minute. This content distribution model offers an additional benefit of fail-over. If one content delivery server to become unavailable for any reason, new requests would continue to be served by the remaining content delivery servers and your user's access would be uninterrupted.

If you allow readers to create user-generated content by creating forms, you must configure a user-generated content server. This server becomes a central repository for all user-generated content across the content delivery group. As readers create comments and other user-generated content on the various content delivery servers, those comments are sent directly to the user-generated content server. When a user requests to view user-generated content, each content delivery server will pass the request on to the user-generated content server, so that all users will be able to see the same, consistent set of user-generated content.