Abstract
In recent years, smart phones and wireless networks became more widespread and the demand for cloud storage services is strong. Cloud storage services gradually replace the hard drives in traditional personal computers and become a new storage media of user’s data. With cloud storage services, people can access their data at any place and at any time as long as network access is available. This convenience has fundamentally changed our daily lives. However, the existing commercial cloud storage services do impose inconveniences and limitations. For instances, service providers may put a limit on the uploading file sizes, or a limit on the storage space. As another example, nearly all services require users to manage their data using software released by the providers. Invariably, these software limit the users to only one cloud storage account, which severely limits the storage space. Another great concern of using commercial cloud storage services is that service providers always keep backups for the sake of reliability. Even after users have deleted their data, there is no guarantee that all corresponding backups are deleted by the service providers. This causes a great concern in data security. We propose an architecture for integration of multiple cloud storage services to solve the problems above. In this architecture, files are stored in file trees locally. Remotely, files are stored in multiple cloud storage services such that all storage limits are not violated. Users can add more cloud storage if more space is desired. In addition, user’s data are encrypted before they are stored in the cloud. This architecture requires a server to manage user’s accounts and to handle file synchronization. In this thesis, we address the design issues and the implementation of a service-oriented server .