A Collaboration Query Strategy Based On P2P Database System

P2P database system,which has high reliability and a wide range of applications. It's suitable for groups, agencies, banks that are geographically dispersed. P2P database system has no central server, each node is uniform. So, there are considerable differences between P2P database system and traditional database system in query processing.There are lots of isomerous and isomorphic databases in the network. The paper focus on the collaboration query basing on isomorphic database system.Comparing with the traditional database, P2P database nodes are much more uniform and dynamic,they can enter and exit system at any time. For doing collaboration query on isomorphic database system, we should pay more attention to issues such as resource routing in networks, node's real-time performance, and transmission delay. Collaboration query takes full advantage of parallel processing efficiency, turns complex query into simple sub-queries, and then executes those sub-queries in parallel. Finally, the node which requests the query combines the results of sub-queries. Collaboration query can significantly improve the reliability and efficiency of query processing.The paper devotes ourselves to do research on the collaboration query, designs and realizes each part using MS.Net: parsing the SQL using Antlr which is a tool for generating compiler; decomposing query into sub-queries which are analogous and unattached; reference to the Contract Net Protocol, establishing a reel-time and steady collaboration networks, and then discussing dynamic priority distribution strategy of sub-queries; researching the results combination, and finally submitting the result to users.