How Caching Is Implemented in ISA Server 2004

When you enable Web caching, you configure ISA Server to store Web objects in its cache. ISA Server includes the following features that optimize cache performance:
1- RAM and disk caching ISA Server allocates RAM for caching popular objects and caches other objects on disk. When caching an object, ISA Server first places an object into the RAM cache and then writes objects to disk. RAM and disk caching help to improve users’ access speed to popular Web sites. By default, ISA Server 2004 uses 10 percent of the RAM on the server to cache Web content.
2- Maintaining the RAM cache in physical memory ISA Server never writes the cached information stored in RAM to the operating system paging file. This optimizes access to the cache stored in RAM.
3- Directory of cached object ISA Server maintains a directory of cached objects in RAM to optimize the process of determining whether the server has an object in its cache.
4- Single cache file ISA Server maintains a single cache file per disk partition to retain cached objects so that gaining access to objects does not use additional system resources that are needed for opening and closing multiple files. The cache file size can be configured for each disk partition. The maximum size of a single cache file is 64 gigabytes (GB). If you require a larger cache store, you can distribute it over different drives.
5- Quick recovery ISA Server quickly rebuilds the directory of cached objects at startup, even after an abnormal termination.
6- Efficient cache updates ISA Server automatically determines which objects to retain in the RAM cache. This decision is based on the likelihood of a user’s again requesting the same object, which is determined by how recently and how frequently an object is accessed.
7- Automatic cleanup ISA Server removes objects that have not been accessed recently or frequently when the disk space that is allocated to the cache approaches capacity.

What Is Web Chaining and Caching?
Some organizations include multiple locations with computers running ISA Server deployed in each location. In this scenario, you can combine caching with Web proxy chaining to optimize caching performance. Web proxy chaining is useful when your organization has multiple branch-office locations, but all Internet requests are routed through one location at the head office. In this scenario, you can install ISA Server in each office and then configure ISA Server at the branch offices to route all Internet requests to the ISA Server computer at the head office.

One of the benefits of using Web chaining is the accumulated caching on ISA Server. If all the servers running ISA Server in the branch offices are configured to forward their requests to the head-office ISA Server, the head-office ISA Server will develop a large cache that contains many requested items. Moreover, the local ISA Server will build up a cache of the most requested items from the branch office. The combination of caching at the local branch office and at the head office increases the chances that the Internet content can be delivered to the client without downloading it again from the Internet.

The following steps how Web proxy chaining works in this branch-office/main-office scenario:
1. The client sends a request for Web content to the Web caching server at the branch office. If the Web caching server at the branch office contains a valid version of the Web content in its cache, it will return the content to the requesting user.
2. If the content requested by the branch-office user is not contained in the branchoffice server’s cache, the request is forwarded to an upstream Web caching server in the Web proxy chain.
3. If the upstream Web caching server has a valid copy of the requested content in cache, the content is returned to the branch-office Web caching server. The branch-office Web caching server places the content in its own Web cache and then returns the content to the branch-office user who requested the content.
4. If the upstream Web caching server at the main office does not contain the requested content in its cache, it will forward the request to the Web server on the Internet. The Internet Web server returns the requested content to the main-office Web caching server. The Web caching server at the main office places the content in cache.
5. The main office returns the content to the branch-office Web caching server. The branch-office Web caching server places the content in its cache.
6. The branch-office Web caching server returns the content from its cache to the requesting user.

Google