Managing Distributed Data Storage System Design And Research

In the recent years, the storage intensity of the memory device increases dramatically, consequently the keep increasing scale of the application programs and the variety of the user data files result in a big boost in the data information contained in the current file system. However, the method and approach to organize and manager these files do not develop adaptively with the speed of the information incensement. The "directory-folder-file" format of the current file system is no longer effective and efficient in managing these large-scale files.The writer take a deep research on the current file system, and provides an improved data storage model "Special Data Storage Model" , or SDS model, for facilitating the storage of different types of the data, including structural or non-structural data.The present article provides a detailed description for the design of the SDS system, and explains the structure of the SDS database engine, SDS layer, and SDS-service layer. The present article further describes the design and implementation of the update detection, compatible operation, update detection and conflict resolve module. The present system achieves the goal by expressing the data information into Units, unit Extents and unit connections, utilizing the search view of SQL, synchronizing the data in different data storage source in the SDS model, tracking the updates in the data, and performing the conflict resolution upon detecting a conflict, provides a efficient data storage for personal user or enterprise users.The present system organizes the data into a model composing by units, extents that are associated by connections, thereby provides a kind of structure data that can be used as common data resource. The present system also supporst the share of the stored information and can provide many services such as synchronization service to the third party application programs.