Research And Implementation Of Remote Data Disaster Tolerance System On Linux

With the rapid development of Information Technology, data has become a key resource related to survival and development of enterprise. Therefore, data disaster tolerance received more and more attention in recent years. The work of this thesis revolves around the research and development of data disaster tolerance system.The thesis thoroughly discusses VFS and Inotify mechanism of Linux kernel. Through improving the frame of Rsync and combining with snapshot technology of LVM, a data disaster tolerance system based on Linux is designed and implemented.Firstly, the author introduces the background and significance of the data disaster toralence system, analyzes the domestic and foreign research status, and describes the necessity for development of remote data disaster tolerance system. Secondly, the theory and technical basis for data disaster tolerance system is described in detail. Thirdly, through improving the frame of Rsync, a design scheme for the RDDT (Remote data disaster tolerance) system is proposed. And on this basis, the function modules of RDDT are divided and designed, also the user interface and job state conversion are given. Fouthly, the functional modules of RDDT are implemented, by using VFS and Inotify mechanism to intercept I/O of system, using snapshot technology of LVM to backup and recover data so as to reduce the resources consumption on the server system, and using FCL (File change log) to locate the changed files in order to reduce the amount of network traffic data and to speed up disaster recovery. Finally, the system is tested, and the running effects of the system are showed.At present, the function of RDDT has been realized initially. Its running shows that the design scheme is reasonable and the expected goals are achieved.