Research On Virtual Machine Scheduling Algorithm And System Implementation In Cloud Computing Environment

With the fast development of the Internet,cloud computing is being used in more and more scenarios in society,and the scale of cloud data centers is gradually expanding,which leads to the cloud data center generated more energy consumption and waste of resources.As virtualization technology is the core technology of cloud computing,it is an important issue in cloud computing research to reduce the energy consumption and improve the resource utilization of data centers through virtual machine scheduling algorithms.To address the above issues,this dissertation investigates two aspects of virtual machine placement and virtual machine migration.The main research elements of this dissertation are as follows.(1)For the study of VM placement algorithms,to address the problem that the existing VM placement algorithm ignores the impact of unplaced VMs on the placement result in the VM placement process,and the impact of VM placement on the overall resource utilization of all physical machines resulting in the algorithm easily falling into a locally optimal solution,a weighted PageRank-based VM placement algorithm is proposed in combination with the PageRank algorithm,the overall resource utilization of unplaced VMs and physical machines are considered,so as to improve the resource utilization of physical machines in cloud data centers and reduce the number of active physical machines.(2)For the study of VM migration algorithms,to address the problem that existing VM migration algorithm ignores data centers cooling costs,heat recirculation effects and local hotspots,a heat recirculation aware multi-objective VM migration algorithm is proposed to consider the three objectives of resource utilization,cooling costs and migration costs,create a multi-objective problem model,and use bi-population evolutionary method for the problem that the NSGA-III algorithm has poor convergence and easily falls into a locally optimal solution.Finally,the improved NSGA-III algorithm is used to calculate the optimal solution set of the multi-objective model and get the optimal migration strategy according to the specific business situation,so as to reduce the cooling and migration costs of the data centers and improve the resource utilization.(3)Design and implement a VM scheduling system according to the proposed VM placement and migration algorithm,which consists of a placement module,a monitoring module and a migration module for the placement,monitoring and migration operations of VMs in the cloud data center.System testing shows that the system can effectively place and migrate VMs.