Research On Virtual Resource Management In Cloud Computing Environment

Cloud computing and virtualization technology have subverted the traditional computing model and become the focus of researches, but there are still many challenges in the study of cloud computing and virtualization, especially in the management of virtual resources. The existing research on the management of virtual resources only considers unilateral factors and ignores the performance parameters which users concerned about. Therefore, study on the technology of virtual resource scheduling and management is of great signif icance and application values in terms of improving resources utilization and utility of the cloud data center and reducing energy consumption.This paper mainly studies the technology of virtual resource management in the cloud environment. To satisfy users’ Qo S requirements while improving the resource utilization, this paper proposes VCRAM- U, a utility maximization model of virtual resource scheduling in cloud environment. In this model, the allocation of virtual cloud resources is abstracted as a utility-maximization problem, which takes tradeoffs between the utility of the data center and the performance of the applications into account, and maximizes the utility on the premise of meeting users’ requirement for performance. A local decision algorithm and a global decision algor ithm are designed to solve the model. Exper iments show that virtual cloud resources can be managed and allocated efficiently by the model and algor ithms. To reduce energy consumption while improving utility of the data center in the cloud computing environment, energy consumption and utility of the data center are regarded as objective functions. VMSA-MOP, a virtual machine scheduling model based on multi-objective optimization is set up and a virtual machine scheduling algorithm based on NSGA-2 is designed to solve the model. Experiments show that compared with other virtual machine scheduling algorithms, our algorithm can obtain optimal scheduling results.Finally, the paper extends the cloud computing simulation platform Cloud Sim and realizes the platform for virtual resource management based on Cloud S im. Exper iments are conducted on the platform to verify models and algorithms proposed in this paper. Experiments show that our models and algorithms have better performance in scheduling virtual resources and can improve resources utilization, increase utility and reduce energy consumption of the cloud data center.