Performance Weighted Deploying and Scheduling Strategy Research for Virtual Machine on Clouds

A performance weighted deploying and scheduling strategy for virtual machine on clouds (PWDSS) is introduced in this paper concerning users´ requests of virtual resources and cloud load balancing. This approach follows three stages: first to use a monitor toolkits to collect the cloud performance data from the virtual machines and physical machines of cloud, and to standardize them, Second, to propose a cloud platform load balancing measurement model, in which the weighted vectors and matrix are set according to the customer requirements, Third, to give an algorithm to select the best appropriate physical machine in the measuring model obtained in stage 2, and then to deploy the new virtual machine, forecasting the load balancing value of every physical machine when the new virtual machine is deployed on it. The experimental results demonstrate that the proposed PWDSS can achieve better effects of system load balancing. At the same time, it can also meet the user requirements better.