Data integration strategy for database grids based on P2P framework

Abstract

The differences between the data integration of a dynamic database grid (DBG) and that of a distributed database system are analyzed, and three kinds of data integration strategies are given on the background of DBG based on Peer to Peer (P2P) framework, including the centralized data integration (CDI) strategy, the distributed data integration (DDI) strategy and the filter-based data integration (FDDI) strategy. CDI calls all the database grid services (DGSs) at a single node, DDI disperses the DGSs to multiple nodes, while FDDI schedules the data integration nodes based on filtering the keywords returned from DGSs. The performance of these three integration strategies are compared with and analyzed by simulation experiments. FDDI is more evident for filtering the keywords with data redundancy increasing. Through the reduction of large amount of data transportation, it effectively shortens the executing time for the task and improves its efficiency.