Research And Implementation Of Cloud Computer System Based On Openstack

With the development of computing scientific and commercial computingsoftware model and architecture increasingly rapid change, while promoting gridcomputing, parallel computing, distributed computing has rapidly developed intocloud computing. Cloud computing, including Infrastructure as a Service (IaaS),Platform as a Service (PaaS), and Software as a Service (SaaS), provide unifiedmanagement and scheduling through these technologies. As a new computing model,cloud computing has been rapidly development by virtue of its low cost, highefficiency, but also to promote the continuous development and improvement of opensource cloud computing infrastructure in recent years.OpenStack is an open source cloud computing projects and tool set, andInfrastructure as a Service (IaaS) solution.The OpenStack not only can quickly deployvirtualization environment, and through this environment to create multipleinterconnected virtual server, and enables users to quickly deploy applications on avirtual machine.In addition, Hadoop as a service (PaaS) open-source distributedcomputing architecture which provided by the Apache-based platform has beensuccessfully applied in Baidu, Amazon, Facebook, and other large sites at home andabroad. The core design of Hadoop are MapReduce and HDFS. MapReduce isdesigned to decompose the task into several sub-tasks to the node processingrespectively, and the results to return to the main node in last. HDFS which namedthe Hadoop Distributed File System, is to provide underlying support for distributedstorage.The focus of this study is using OpenStack to build a cloud computingarchitecture which is based on Infrastructure as a Service(IaaS), and on which we willbuild a Hadoop environment, according to this structure to achieve the integration ofthe Infrastructure as a Service(IaaS) and Platform as a Service(PaaS), and provideuser interface of cloud computing with providing user a web page. The core idea ofthis architecture is to use OpenStack to build any number of private clouds, each ofthe private cloud is a service node and each node has multiple resource nodes, anddifferent private cloud and the resources of private cloud are independent with eachother. According to OpenStack to complete the configuration and deployment of the service nodes and resource nodes,and at the same time to deploy Hadoop in eachresource node.User can submit their own task when they apply the resources, cloudcomputing platform can assigned to different nodes randomly based on the resourcesand will return the task results after completing the calculation.The architecture of this thesis is divided into three layers: the infrastructure layer,the middleware layer and the portal layer of cloud.The infrastructure layer is made ofvirtualization technology Xen virtualization tools and the OpenStack structures.Middleware layer and portal layer is built on the infrastructure layer. Foremost amongwhich is the middleware layer, which encapsulates the communication module, filetransfer module, the remote object module. Communication module complete thecreation and deployment of cloud service nodes and resource nodes, achieving theallocation of computing resources and storage, it uses WebService provides a unifiedinterface for the upper and future secondary development. The file transfer moduleusing FTP technology achieve real-time interaction and feedback of the server andclient. Remote object module to complete the release of the application layer, anddemonstrate the result more clearly and quickly in various forms.When the thesis study and design the cloud computing platform, it considers thetastes of user and provides user with the Infrastructure as a Service (IaaS) functionthrough the Web. Cloud user can submit tasks through the Web page and applicationresources,while managers can devote their effort to resources for effectivemanagement and deployment.