National University of Defense Technology
Infrastructure-as-a-Service (IaaS) clouds provide on-demand Virtual Machines (VMs) to users. How to improve the quality of IaaS cloud services is important for service providers. Currently, the VMs in an IaaS cloud are usually deployed with respect to the maximum utilization of resources. In this paper, the authors propose an online VM optimization method for IaaS clouds. Their method mainly optimizes the VM deployment in IaaS clouds according to the traffics among VMs. VMs are allocated with respect to cabinet capacities at the beginning.