Updating cache content automatically
Microsoft Internet Security and Acceleration (ISA) Server extends caching performance further with customizable cache download. By monitoring and analyzing Internet access, you can determine which Hypertext Transfer Protocol (HTTP) content will be needed and when. You can then use the scheduled cache content download feature to download the HTTP content directly to the ISA Server cache, upon request or as scheduled. This enables you to update the ISA Server cache with HTTP content that you anticipate will be requested by clients in your organization. This content will be available for access directly from the ISA Server cache, rather than from the Internet.
You can download a single Uniform Resource Locator (URL), multiple URLs, or an entire Web site. When you create a scheduled cache content download job, you can limit which content should be downloaded. For example, you can limit the download to a single domain. Similarly, you can limit how many links to follow. For more information, see Configure a location from which to download content.
When you schedule cache content download, you can configure dynamic content to be cached by configuring the ISA Server cache to store the objects, even if the HTTP cache control headers indicate that they are not necessarily cachable. For configuration instructions, see Cache dynamic content for scheduled content download jobs.
Warning
| • | If the Web server from which the object is being downloaded requires client authentication, then the scheduled content download job will fail. |
The download occurs according to a preconfigured schedule. (You can also specify whether the schedule should recur periodically.) Scheduled content download jobs can be configured for outgoing Web requests and for incoming Web requests.
For outgoing Web requests, you determine which objects on the Internet are most often requested by users. You can then schedule jobs that retrieve the objects from the Internet and load them into the cache.
For incoming Web requests, you can schedule jobs that retrieve content from your internal Web servers.
For more information, see Configure content download job schedule.
Cache Array Routing Protocol and scheduled content download
When Cache Array Routing Protocol (CARP) is enabled, ISA Server uses the CARP algorithm to determine on which server the object should be cached. When CARP is disabled, the object is cached to all the servers in the array. For more background information, see Cache Array Routing Protocol.
By default, CARP is enabled for outgoing Web requests and disabled for incoming Web requests. That is, by default, ISA Server uses CARP to cache objects from outgoing Web requests on just one specific server, but objects from incoming Web requests are cached on all servers in the array.
ISA Server checks the outgoing Web request settings for CARP when retrieving objects for a scheduled content download job. This is true for download jobs that apply to outgoing Web requests and to jobs that apply to incoming Web requests. For this reason, it is recommended that you disable CARP for outgoing Web requests, if you want the scheduled content download job to cache objects to all the servers in the array.
