The Design And Implementation Of A Cloud Storage Service Based File System

With the development and application of information technology in different industries and fields, many organizations, institutions have accumulated more and more industry data. Belonging to the explosion of data, the requirements of data computation and storage also presents the increasing trend. The concept of cloud computing arises at the historic moment. Many organizations participate in the research and commercialization process. As a result, cloud computing obtained the huge development. But it brings some problems, such as repetitive construction and investment. Different industries and areas are busy in building cloud computing data center.Proposed the concept of cloud computing platform was originally intended to integrate all available resources, namely combine the computing resources of hardware together as resource output, rather than extravagance and waste of resources. There are many companies or organizations that have opened there cloud computing resources, but the use of these resources is need to pay. Although there are some products provided free strategy, but its capacity of cloud storage resources or access mode have certain restrictions.This paper explore how to make cloud storage platform construction as far as possible to reduce the input of local hardware resources and make full use of open cloud storage resources in order to achieve good performance. Based on the studying knowledge of the file system infrastructure, existing distributed file system and cloud file system, this paper put forward a design scheme and finally implement the system. This designed file system transfer underlying data to cloud storage service. The file system’s requirement of local hardware equipment is low, generally it only need one server. Besides, the file system have a good access performance and management. At the same time it provides a friendly operation interface to the users.In this paper, the main research work and achievements are as following three parts:(1)The system design. On the study of the traditional file system architecture and analysis of the related cloud file system design idea and scheme. A new file system design scheme of storing underlying data to cloud storage service was brought out.(2)The system implementation. This paper implements the file system’s overall architecture and basic file operations function module. Through the metadata and data block separation management design, as well as the design of metadata management, this paper realize the synchronization function between different user terminals and cloud data.(3)The system optimization. By distinguishing file type to store data, the performance of file system on read and write functions has been promoted. What’s more, uid prefetching technique improves the data access speed, so as to improve the system read and write performance.File system’s overall performance and the performance of two optimization are in laboratory tests have carried on the qualitative and quantitative comparison experiment. Experimental data clearly reflect the overall performance of the file system. It is consistent with the expected design. Optimization design significantly improve the performance of the file system on read and write functions.