Font Size: a A A

Research And Implementation On Integrating P2P Technology Into Fast Content Distribution Mechanism In Cloud Platform

Posted on:2017-09-13Degree:MasterType:Thesis
Country:ChinaCandidate:W J ZhaoFull Text:PDF
GTID:2348330485961580Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
With the development of cloud computing, the cloud storage services have became universally applicable storage solutions for business applications and personal applications. When users make requests to download files from the cloud storage server, these servers begin to provide content distribution services. During the content distribution process for data transmission, HTTP protocol is usually used. However, there are some problems about such content delivery process. A large number of users need to download the same file from the cloud storage server in a very short time, so it makes big pressure onto the cloud server bandwidth, and makes the client downloading more slowly. To solve this problem, we propose a rapid content distribution mechanism based on OpenStack cloud platform which integrates P2P technology. This integrated content distribution method makes use of advantages of sharing data among P2P client to reduce bandwidth consumption in cloud server side and improve the download speed in client side.The major work in this thesis concludes three related aspects. First, we propose several reference metrics and their calculation method for rapid content distribution mechanism. In order to determine the exact condition of protocol conversion between HTTP and P2P protocols, and show effects of rapid distribution mechanism, we select four metrics:user type, quality of service, time gains, the bandwidth revenue. The time gains and quality of service mainly focus on the interests of users, and the bandwidth revenue and the user type mainly focus on the interests of the cloud service provider. We show the calculation method for time gain metric and bandwidth revenue metric precisely, and make adaptive revision about the small file distribution. Second, based on these four measures metrics, we propose five feasible dynamic protocol conversion methods. Through performance analysis of each method, we choose one protocol conversion method as the most suitable one for file sharing scenario. Based on this method, we calculate the bandwidth threshold which can be redistributed for following appropriate bandwidth resource scheduling. Finally, we implement above protocol conversion method based on OpenStack cloud platform, and discuss the key technologies.Through the experiment, we validate and illustrate advantages of our proposed dynamic protocol conversion method, that is, our method could effectively reduce the user download time and save the bandwidth resources of the service providers.
Keywords/Search Tags:Content Distribution, BitTorrent protocol, Reference metrics, OpenStack, Cloud storage
PDF Full Text Request
Related items