Peer-to-Peer (P2P) is an architecture in which participating machines can both receive and provide data or resources. Each participant is called a peer.
For example, a machine downloading a file can send the parts it already has to other machines. Data moves between participants instead of every download coming from one central server.
A P2P system can still use servers to help participants join the network or find peers. Those servers do not necessarily carry all the transferred data.
Suppose an application splits a file into several parts. Your machine finds peers holding the missing parts and can download different parts from several peers at once.
After receiving a part, the application compares its hash, a value calculated from the content, with the expected value in file metadata obtained from a trusted source. A valid part is retained and can be shared with others. Once all parts arrive, the application assembles the complete file.
If a peer disconnects, the application looks for another peer holding the same part. Downloading can continue only if that part is still available somewhere; participating in P2P does not automatically keep a file available.
P2P suits file sharing, content distribution, and applications where participants exchange resources. The system must handle peers joining and leaving and provide a way to retain important data.
For business websites requiring centralized administration and close control of data, a client–server design is often easier to build and operate.
A library of IT interview questions with detailed answers — from Junior to Senior.
Donate