The Design And Implementation Of Cloud Synchronization System Based On Cloud Storage

As the high development of Internet and IC technology, users can store and manage data in different intelligent terminal devices. To prevent the loss of data, user can store backup data in safe places. Data synchronization between the terminal devices is also convenient for user to access data on different terminal devices. To meet demands of data backup and synchronization, user often choose synchronization software to manage data. Benefited from the rapid development of cloud storage technology, the issue of large-scale data storage in synchronization software also has been resolved. The main work of this thesis is as follows:This thesis designs and implements a cloud synchronization system based on cloud storage, and studies the design and corresponding algorithms of the system. It first introduces the design of each module, describes the system architecture and access interfaces of the cloud storage platform based on OpenStack Swift. Respectively, execution mode, role of each part and database design of server and client are described and analyzed. Then according to characteristics of version, the version control model of the system and implementation algorithm of version control strategy is described. Finally it elaborates the operation mechanism and synchronization strategy of the system.In order to improve the utilization of storage space of cloud synchronization system, this thesis focuses on the data storage strategy. And on the issues of ver-sion storage of large file, a new incremental storage algorithm is proposed. The algorithm is based on inverse incremental storage model, using the incremental calculation method of the matching block checksum. As a result, storage space of the system is greatly reduced. Furthermore, this thesis studies algorithms of version storage and recovery function based on the incremental storage algorithm. Through the comparison with several existing storage algorithms in experiments, the proposed incremental storage algorithm has been proved efficient and avail-able.