Data Consistency Research In P2P Environment

With the rise of technology, peer-to-peer (P2P) computing is an important network computing in recent years.There are a large number of researches and applications in many areas. With the continuous development of P2P network, users have a higher request with the effectiveness, real-time of P2P network resources and transmission efficiency. In P2P networks, data replication is a key technology in improving the resource availability, reliability and transmission efficiency in the distributed system. However, the traditional way of creating a copy is a static. With the improvement of user requirements for resources, to research how to effectively manage data consistency for the replicas’resources which is updated dynamically is very important. This consistency is not only to ensure the strong consistency but also to satisfy the requirements of consistency updating in a fast speed.This paper describes the characteristics of P2P and four of P2P network topology structure in detail. According to a copy of the P2P network environment being likely to cause data inconsistencies easily under the data replication and caching, proposed four standards of data consistencies:strict consistency, number sequence consistency, weak consistency and eventual consistency. On the basis of having a deep understanding of the P2P network topology structure and creating replicas in the P2P environment, the data update algorithm are put forward under the environment of structured P2P system by analyzing and comparison to the current existing problems, and combining the data consistency updating technology in P2P environment:The consistency updating algorithm based on Chord protocol of d-fork propagation tree(hereinafter referred to as dDT). The algorithm maintains the consistency between the node and it’s replica by the establishment of d-fork propagation tree, furthermore, it assigns a m bit identifier keywords for each node separately through designing the consistent hash function, then divide into the Chord ring composed of nodes,which can spread the update information more quickly so that shorten the update propagation delay. More ever, the algorithm deal with the failure nodes of the network by the tree node switching strategy, thereby reducing the discard rate update. By comparing with other methods of consistency maintenance, it is to illustrate the algorithm’s superiority in redundancy updates, node coverage, update propagation delay,etc.Finally, the simulation analysis through the P2P simulation tools from redundant information, extensibility, a node updates operation with the number of messages, the algorithm of overhead updating time delay spread, etc is used to the performance analysis, which is to demonstrate the algorithm in P2P dynamic environments feasibility of maintaining data consistency. There are21figures,3tables and57references.