P2P(PEER to PEER) Technology <ul><li>P2P is a Networking Software where clients communicate with each other rather than go through a server </li></ul><ul><li>The most common way of linking computers so they can share data or processing power is via client-server architecture. Servers store data and applications that PCs can access. </li></ul><ul><li>A different way to achieve data and application sharing is by allowing PCs to connect directly with each other and act as both clients and servers. This is the basis of Peer-to-Peer computing. </li></ul><ul><li>A number of years ago applications that allow these direct connections were developed for use in office environments with a small number of PCs. </li></ul><ul><li>Programs such as AppleTalk Network and Windows for Workgroups are early examples of this technology known as Peer-to-Peer (P2P). </li></ul>
P2P continued……… <ul><li>Recent developments have made it possible to expand P2P technology to computers connected to the Internet as well as those connected through a network. </li></ul><ul><li>P2P programs fall into two major categories: </li></ul><ul><li>(i) centralized </li></ul><ul><li>(ii) decentralized . </li></ul>Centralized System In a centralized system a server maintains directories of what is stored in each peer computer. The central server directs the connections between peers. Thus, while the connection is made directly from peer to peer, a server is still necessary
Server User User User User User User User User Search Request Search Response Download Centralized Model
P2P continued…… Decentralized System With a decentralized program each peer connect directly to other peers without being directed by a server. These connections are repeated making the resulting network, in theory, infinitely large. In reality this size of a network is constrained by the number of layers of computers a message has to travel through.
Search Request Search Relay Search Response Download User User User User User User User User User User Decentralized Model
<ul><li>History of Napster </li></ul><ul><li>Two college students, 19-year-old Shawn Fanning and 20-year-old Sean Parker founded Napster in late 1999. </li></ul><ul><li>They originally just wanted to trade music over the Internet with their friends, but their P2P network has since grown to a user group of over twenty million. </li></ul><ul><li>The Napster phenomenon has increased interest in peer-to-peer networks to a fever pitch . </li></ul>Boston Students !!
Napster continued…. Napster is a centralized or server-based P2P network. With a server-based network, there is a central list of information that is accessed by all the users of the system. Here is what happened when you decided to look for the song: 1.You opened the Napster utility. 2.Napster checked for an Internet connection. 3.If it found a connection, Napster logged you onto the central server. The main purpose of this central server was to keep an index of all the Napster users currently online and connect them to each other. It did not contain any of the MP3 files. 4.You typed in the title or artist of the song you were looking for. 5.The Napster utility on your computer queried the index server for other Napster computers online that had the song you requested.
Napster Continued…. 6.Whenever a match was found, the Napster server informed your computer where to find the requested file. 7.When the server replied, Napster built a list of these systems in the results window. 8.You clicked on the file(s) that interested you and then chose Download . 9.Your copy of Napster attempted to establish a connection with the system hosting the file you selected. 10.If a connection wa successfully made, the file began downloading. 11.Once the file wa downloaded, the host computer broke the connection with your system. 12.You opened up your MP3 player software and listened to the song.
Comparisons of Centralized and Distributed Model <ul><li>File is instantly Available in Centralized Model but in Distributed Model it is subjected to Availability </li></ul><ul><li>Shutdown of Distributed Model is Impossible but is possible to turn off the Central Server </li></ul><ul><li>There is no Bootstrap Problem in Centralized model </li></ul><ul><li>Flooding of request packets occur in Distributed Model but not in the case of Centralized Model </li></ul><ul><li>Search time is higher in Distributed Model </li></ul>
Points to Ponder…. <ul><li>P2P is Good or Bad ? </li></ul><ul><li>P2P eats up all the Bandwidth, As an Engineer what we are going to do to meet the Bandwidth requirements in the Feature.? </li></ul><ul><li>What is the Difference Between SOA and P2P ? </li></ul><ul><li>What AI Techniques can be applied to p2p to make it Better in the future ? </li></ul><ul><li>Do we have any ideas to stop piracy in the future ? </li></ul>
Hybrid P2P Technology <ul><li>The Hybrid P2P is a combination of the best features of peer-to-peer delivery together with the best features of central server or unicast delivery. </li></ul><ul><li>We have taken the security, high quality, and control of unicast technology and combined it with the extreme efficiency of peer-to-peer delivery. </li></ul><ul><li>The result is a very secure, high quality, stable, quickly adaptable network that uses up to 98% less bandwidth. </li></ul><ul><li>This makes it better than unicast and better than pure peer-to-peer. </li></ul><ul><li>It also creates proven higher quality connections. </li></ul>
Research on P2P… <ul><li>Research can be Done on </li></ul><ul><li>Architectures </li></ul><ul><li>Algorithms </li></ul><ul><li>Performance Modeling and Optimization </li></ul><ul><li>Distributed Resource Management </li></ul><ul><li>Privacy, Security, Trust and Incentives </li></ul><ul><li>Applications </li></ul>
<ul><li>Topology Building of P2P Networks </li></ul><ul><li>Game theory Random Routing </li></ul><ul><li>QoS Routing </li></ul><ul><li>Updating Routing Table by Sampling </li></ul><ul><li>Hybrid Information Retrieval </li></ul><ul><li>Ant Colony Algorithm </li></ul><ul><li>Semantic P2P Changing Node Join Process </li></ul><ul><li>Topology Building of P2P Networks P2P Routing Optimization </li></ul><ul><li>Using Detector to update connectivity </li></ul><ul><li>Key Words Partial Matching Search in DHT Routing </li></ul><ul><li>Adding New Data Structure in Routing Table to change searching way </li></ul><ul><li>Peer Node </li></ul>Research cloud….
Common Terms Used in P2P Peer A peer is anyone who is downloading or uploading the file Client The software that allows you to download and upload (share) files using Torrent. The Torrent client opens the .torrent file and allows connections from your peers. Tracker A tracker is a server that keeps track of which seeds and peers are in the swarm. Clients report information to the tracker periodically and in exchange receive information about other clients to which they can connect Seeder A seeder is a peer that has a complete copy of the torrent and still offers it for upload. The more seeders there are, the better the chances are for completion of the file. Leech A leech is usually a peer who has a negative effect on the swarm by having a very poor share ratio, however it can also refer to any peer without 100% of the data Swarm Together, all peers (including seeders) sharing a torrent are called a swarm.