Research And Optimization Of The Deployment Process Of Virtual Machines Based On OpenStack

Cloud computing is a new computing model,the concept of which evolved from grid computing,distributed computing and parallel computing.In combination with the Internet,it provides virtualized IT resources to customers according to their needs dynamically.Currently,the main resources in cloud computing environment are virtual machines.By the use of virtualization technology,the providers could virtualize hardware resources in data centers.Once the virtual resource pool produced,the providers could allocate virtual machines to customers dynamically by using the resource scheduling technology.This paper analyzes the current status of virtual machine resource scheduling in cloud computing environment and intensively studies the virtualization and resource management technology.With the help of OpenStack open source cloud computing platform,this paper proposes the optimizational technology of virtual machine deployment process based on it,which includes the user filtering mechanisms and the physical machine filtering mechanisms.Only in this way can it ensures the rational usage of physical resources and improves the energy efficiency.This paper puts forward Max-RAM algorithm based on pre-classification to aim at the problem of low utilization rate of resources,which classifies the physical resources by grade through the computing to avoid some of the resources in the resource scheduling congestion or idle phenomenons.Experimental results show that the Max-RAM based on pre-classification algorithm not only avoids partly physical resource congestion or idle phenomenon appearing during the virtual machine deployment,but also improves the probability of virtual machine successfully starting-up.More importantly,it improves the utilization of resources,achieving the purpose of rational use of physical resources.This paper has a good theoretical and practical value.