| With the rapid development of China’s e-commerce, online financial and other information industry, many companies have increasingly recognized the importance on business continuity of information system. As a system to ensure reliable and stable operation of the enterprise information system, Disaster recovery system plays an important role and can ensure that information systems in the event of a disaster will resume normal business service in a short time. In disaster recovery technology, comparison of data consistency can compare data between production center and disaster recovery center after data synchronization is complete and then position variance data so that it can be repaired subsequently to ensure data consistency at both ends. It is important for data integrity and consistency and is an important assistive technical means for failover and failback in disaster recovery system.This thesis aims to design and implement a management platform for centralized management of enterprise’s disaster recovery, and focus on implementing a subsystem for comparison of data consistency. This thesis’ main work is as follows:(1) Study and research theories and technologies about disaster recovery. The related development technologies for this system, including J2 EE development platform and MVC model, are introduced.(2) Based on the demand of disaster recovery for a company’s production system, the overall structure of the disaster recovery system is analyzed from three aspects in architecture, function structures and system messages. Subsystems are divided and the controlled application side is analyzed.(3) The subsystem for data comparison is divided into three modules to design and implement. Data comparison Master Control Module is responsible for implementing the data comparison tasks and return the results. Data comparison Service Module is responsible for data statistics and report it to Data comparison Master Control Module. The delivery of messages between two modules is through Tuxedo messaging middleware and data through TCP/IP. The policy configuration and execution module is responsible for configuring parameters and initiating a data compare comparison task. This thesis detailed analysis data comparison algorithms and selects one appropriate algorithm for the subsystem.Through test and analysis, the data comparison subsystem used for this disaster recovery management platform can achieve the specific function of data comparison between the primary and secondary ends and the performance for data comparing can meet the requirements. This subsystem can be efficient, reliable and stable operating. |