Research On Separation Of User Data And Operating System In Desktop Virtualization

With the development of cloud computingtechnology,desktop virtualizationhas widespread application in the field of education, health care, finance, life.Giant IT companieshave developed a series of desktopvirtualizationproductsat home and abroad.Compared with the conventional computers,the use of desktop virtualization reduces the energy consumption, makes centralized management easier,improves the data security and enhances the user experience degrees. However, system image file includes operating system, applications, and user data.With the continuous growth of user data, the storage space of desktop image file increases. It brings great pressure to the storage and management of image file in the server.After analyzing domestic and foreign current research status, according to the currentuse status of desktop virtualization, this article uses the way separating user data from the operating system to reduce the size of desktop image file.Based on the research purpose, the main work of this paper includes:expound of the development process, opportunities and challengesabout desktopvirtualization,propose of problem about the increased costs of management and storage of desktop image file because of the continuous growth of user data;analyzing related workabout theprocessing of file data in desktopvirtualization which domestic and foreign researchers have done;usingFileSystemWatcher classprovided by C#to realize the monitoring and analysis of file system operation behavior; accurate classification foruser data though excluding the interferencedata of the application and system operation; formulation of different data synchronization strategies to realize the synchronization of user data from client-side to server-side; the use of different data transmission technology to realize the security of data transmission;check the consistency of the client and server data,realization ofseparation of the client’s data.This paper has tested function of important program and efficiency of the algorithm. The test results meet the research and realize research purpose.Finally, according to the existing research results,the next work is prospected which includes four aspects:the improvement of the performance in processing large amounts of data of file system monitorprogram, getting the user file data from special path, the optimization of Rsync algorithm and how to access server-side data.