Research And Implemention Data Replication Between Remote Databases With Multithread And Continuous Transmission On The BreakPoint

Various operation systems based on database have been established by companies and institutions. These operation systems have acclumated a large amount of precious data, the data have been provided basis for more advanced analysis. In order to take full use of precious data and get efficient information from these data, it needs to send the seperated data to centralized database or data warehouse for analysis. Now available data transport technology need retransfor the data when network unstable or interruption. The data size of replication may be very large, meanwhile the traditional form of data transform is iterator of request——transform—request—transform, and this form can not take full use network resource. At last, there is no standard method to solve the constraint difference between heterogeneous databases.The completed work and innovation of this paper are as follows:At first, consider the complicated constraints of relational database, designed an extended topological sorting, the new arithmetric can deal with dependency with each other and circle dependency.Secondly, the amount of data replication may be very large, designed a new JDBC data trnasfer model. In the new JDBC model, several network connections can transfer the same result set parallelly, and it can increase the efficiency of network.At last, based on the new JDBC driver, this subject designed a data replication system. It realizes data replication between heterogeneous databases and it can realize multithread transfer and continue transfer on the breakpoint when the network reconnects.