The Study And Construction For The Virtualization Platform Of Campus Based On Cloud

With the development of cloud computing, the expansion of enterprise data and application scale, the management and maintenance of data center has become increasingly complex. How to improve the utilization and sharing rate of data center resource and enhance system stability has become an important goal of information construction in universities. How to improve the utilization and sharing rate of data center resource and enhance system’ stability has become an important goal of information construction in universities. By researching the theory and technology of virtualized cloud environment, the utilization of data center resources is improved, planned downtime is eliminated, and system management is simplified. Additionally, by focusing on the virtual machine deployment and migration strategies of the server virtualization, the load balancing of data centers is achieved and the resource utilization is improved.In this paper, the analysis, research and practice of server virtualization, storage virtualization, network virtualization and desktop virtualization’ purpose, key technologies, virtual models and the deployment method are researched. The purpose is to provide reference for users who want to build a virtual campus. Server virtualization is mainly to optimize the allocation of resources, improve the resource utilization, and increase the availability and continuity of systems. Storage virtualization is to achieve a unified data management, synchronous backup and disaster recovery, which can ensure consistency and security of data, provide efficient data access services of systems running and achieve storage system’ dynamical expansion. Network virtualization is to optimize network structure, improve network bandwidth and traffic, and ensure network security and availability, which can achieve full utilization of network resources. Desktop virtualization is to further enhance the user’s desktop effects, to satisfy any walk in the cloud of user’s desktop, which can achieve supercomputing and storage capacity of the client.For the issues of server virtualization’ load balancing and resource utilization, we design a virtual machine deployment and scheduling methods of optimized AHP. Firstly, according to the classification of the virtual machine, and by using the server’ avoiding mechanism, the target server is screened preliminarily. Secondly, the weight vector of the target server selection is determined by using AHP, which includes the virtual machine memory, CPU and bandwidth. Finally, according to the product of the target server resource usage and the weight vector, the virtual machine server is predicted and evaluated. When deploy the virtual machine, the best scheme is chose by estimated value. When migrate a virtual machine, the server load is predicted and evaluated, and according to relative ease degree of the target server and load server, the best scheme of load balancing is chose. The experimental results show that the strategy can not only improve the server load balancing and resource utilization, but also can optimize the allocation of resources.