Research And Implementation Of Resource Allocation Based On Openstack Business Cloud Platform

Cloud computing has caused a great revolution in IT industry. It not only changes the way for IT companies to deliver their softwares/hardwares, but also affects how customers will store and access information over the Internet. Many service providers have decided to deploy their services on cloud platform to enjoy the great benefits of cloud computing. The efficiency of resource allocation on business cloud platform will directly affect the performance and cost of services, thus this is a key problem to solve during the commercialization process of business cloud platform. Therefore, the research on resource allocation on business cloud has great academic and practical value.We learn the basic knowledge and key technology related to cloud computing and make a comparison of the several common open source IaaS cloud platforms. According to the characteristics of the platforms and our needs, we decide to build a private business cloud based on OpenStack. Then we deploy the multimedia sharing service on the cloud. After the study of several current resource allocation algorithms in cloud computing environment, we carry on the analysis and modelling of the allocation problem on the cloud platform. Then we propose and implement a resource allocation model based on the characteristics of business cloud platform. In this model, the resource allocation is performed at two levels:at the host level and at the virtual machine level. At the host level, we adopt the simple allocation algorithm inside OpenStack to allocate the physical resources to each virtual machine. At the virtual machine level, we model the problem as a variable-sized bin packing problem and propose an algorithm to assign the resources to each service component.Experimental results indicate that the resource allocation model works well on the business cloud platform. The allocation solution can assure the service performance as well as decrease some additional overhead, making the use of resources more effectively.