Font Size: a A A

Design And Implementation Of Infrastructure Management System In Cloud Computing Center

Posted on:2013-09-28Degree:MasterType:Thesis
Country:ChinaCandidate:F JiFull Text:PDF
GTID:2248330392453224Subject:Software engineering
Abstract/Summary:PDF Full Text Request
With the benefits of cloud computing model and virtualization technology, thispaper assumes that the next generation of Internet computing center will be aco-existence state with multiple types of virtual clusters, involving the traditionalhigh-performance computing clusters, physical or virtual clusters for large-scale dataprocessing in MapReduce mode and virtual clusters for multi-tenant usage.Meanwhile, in order to provide computing power in a wider geographical range, anumber of Internet Computing Centers will collaborate with each other to share moreresources. In such a complex infrastructure usage pattern, it is a big challenge toachieve effective management of large-scale infrastructure and on-demand computingcapacity for applications.This article addresses this issue, outlines three common usage scenarios from thepractical experience and summarizes their various requirements on infrastructuremanagement platform. This paper combines with these requirements, proposesprototype system design principles and programs for the next-generationinfrastructure management in Internet High Performance Computing Centers, andshows specific routes of system implementation. After studied the traditional highperformance computing, grid computing and cloud computing with virtualizationfeatures, we introduces a cloud computing platform, which is derived from Publicsecurity system in the cloud computing center project.The platform support the traditional high performance computing supportMapReduce large-scale data processing and support the coexistence state of physicaland virtual fleet of multi-tenant usage patterns. To a wider range of computing power,our system provides each other with more resources among multiple InternetComputing Center through the collaborative methods.This paper clarifies the whole system needs analysis, the overall architecturedesign, detailed design, system testing, system deployment. The entire project tookover two years, and ultimately achieved very good results. By walk-through approachof scenarios, a simple assessment is given for the system design. Finally, the papersummarizes the possible key technologies and innovation in the future.
Keywords/Search Tags:Cloud Computing, IAAS, MapReduce, Grid Computing
PDF Full Text Request
Related items