1.3.1 Peer-to-Peer Framework Details

  • Article

BITS discovers peer servers by using the Background Intelligent Transfer Service (BITS) Peer-Caching: Peer Discovery Protocol specified in [MS-BPDP]) and authenticates them by using the Background Intelligent Transfer Service (BITS) Peer-Caching: Peer Authentication Protocol specified in [MS-BPAU]).<2> For more information on BITS, see [MSDN-BITS].

The BITS Peer-Caching: Content Retrieval Protocol does not address issues of cache management, such as policies for adding and removing content or the method of storing and indexing the content. These issues are internal to the server implementation.

The following figure shows a black-box diagram of the BITS peer-to-peer framework.

Black-box diagram of BITS peer-to-peer framework

Figure 1: Black-box diagram of BITS peer-to-peer framework

To start, the client initializes the BITS Peer-Caching: Peer Discovery Protocol client role to listen for hosts that support the server role of BITS Peer-Caching: Content Retrieval Protocol (for more information, see [MS-BPDP] section 3.2.3).

When a BITS Peer-Caching: Peer Discovery Protocol server is initialized, it announces its presence to clients as described in [MS-BPDP]. Thus, over time, clients gather a list of nearby peer servers.

Because the BITS Peer-Caching: Content Retrieval Protocol implements a cache, the client does not search for data or download it until requested by a higher-layer protocol. A typical usage pattern is as follows:

  1. External to the BITS Peer-Caching: Content Retrieval Protocol, a client application identifies the need to download a particular URL, with a known timestamp and length.

  2. The application initiates a BITS Peer-Caching: Content Retrieval Protocol search to determine whether any peer servers contain the necessary URL data. The BITS Peer-Caching: Content Retrieval Protocol client chooses a set of peer servers and queries them for the URL data. BITS Peer-Caching: Content Retrieval Protocol clients and servers use the BITS Peer-Caching: Peer Authentication Protocol to verify that the peer is a member of the same Active Directory domain, as described in [MS-BPAU].

  3. Based on the success or failure of the search, the application downloads the URL data, using BITS Peer-Caching: Content Retrieval Protocol if possible or HTTP(S) from the origin server if not.

  4. If the client host also implements the server role of BITS Peer-Caching: Content Retrieval Protocol, then the application tells the BITS Peer-Caching: Content Retrieval Protocol server role to add the URL data to its cache, thus making it available to other peers.

For more detailed sequence diagrams, see section 4.2 and section 4.3.