Linux Platform For P2P Traffic Control Strategy

Today, as the rapid development of the Internet,more and more industries and applications rely on the Internet, Meanwhile, amounts of P2P applications such as BT, eMule etc., are used for file sharing, video publishing/downloading, online gaming, cooperative working, etc. However, based on the principle of Peer to Peer (P2P), network bandwidth will be greatly consumed. The abuse of P2P software rusluts in business-critical applications can not run smoothly. On the other way, as a result of P2P technology fast development, P2P software use various of technologies such as dynamic ports, HTTP carrying P2P data, etc, which make very difficult for firewalls, routers and other filtering device to discover. Therefore, the classification, identification and control of P2P traffic become more and more important.According to these problems, focusing on analysis and research of the current mainstream P2P traffic identification and control technology, was studied under the Linux platform, P2P traffic monitoring of a comprehensive strategy and a detailed design and simulation implementation.Analysis of the structure and working principle of netfilter, the specific understanding of the working principle and configuration of the TC method. Analysis of the TC module, with a specific example describes the whole process of flow control; of traffic based on application layer identification strategy, design flow of a two-stage recognition algorithm, namely, the combined port and load the contents of P2P traffic identification algorithm, the improved algorithm can effectively identify the P2P traffic, given the specific deployment strategy, system architecture diagram and the specific implementation. This flow control strategy greatly facilitated our management of network traffic, the test proved that the module is better able to identify P2P traffic.This paper has carried on the performance test to the P2P current capacity management system, mainly including three aspect tests:TC flow control ability test, application layer recognition module testing, integrated management ability test. From the statistical data analysis, the system is turn to be capable to the P2P application identificaiton, and the control effect is remarkable.