Distributed cache mode example
Updated: November 4, 2009
Applies To: Windows Server 2008 R2
In this example, BranchCache is deployed in distributed cache mode at a main office and a single branch office. This example depicts the following two stage process:
The branch office client computer Client 1 seeks and obtains content from the main office content server.
Another client computer in the same branch office, Client 2, seeks the same content that was originally obtained by Client 1. In this case, Client 2 obtains the content from Client 1 instead of over the wide area network link from the main office content server.
At the outset of the process, the employee using the computer Client 1 attempts to access content that is stored on a BranchCache-enabled content server.
Note
The content that the employee wants to access can be on any of the three types of content servers – a Web server, a BITS-based application server, or a file server.
In the following illustration, Client 1 seeks and obtains content from the main office content server.
The numbers in the illustration above correspond to the following actions:
Client 1 sends a request for content to the main office content server. In this request, Client 1 indicates that it is BranchCache-capable.
The content server obtains previously generated content information from a local cache and sends it to Client 1.
Client 1 uses the content information and sends a multicast message to all computers on the subnet requesting the content; no computers have the content, however, because none of them has previously downloaded the content from the main office.
Client 1 requests the content from the main office content server.
Client 1 receives content from the content server and stores the content in its cache.
In the following illustration, Client 2 seeks and obtains content from Client 1.
The numbers in the illustration above correspond to the following actions:
Client 2 sends a request for content to the main office content server. In this case, Client 2 seeks the same content that Client 1 has already obtained.
The content server obtains previously generated content information from a local cache and sends it to Client 2.
Client 2 uses the content information and sends a multicast message to determine if any clients in the branch office have already cached the content. Client 1 sends a response stating that it has the content.
Client 2 requests the content from Client 1, connects to Client 1, and downloads the content.