Designing automatic power saving on virtualization environment

Currently, the data center is equipped with a huge number of servers, storage and networking devices. Aided by the virtualization technology, the population of servers in the data center can be dramatically reduced, thus enabling a very significant reduction in the power consumption, reduced operational costs, and also enhanced corporate competitiveness. Built upon the virtualization technology and the resource allocation, therefore, this paper will utilize the pre-defined scenario model to propose the scenario-based APS technology. Meanwhile, this research can also integrate the Linux platform to develop the user interface for the scenario-based APS. Furthermore, the virtual framework of the Open Source Hypervisor was set as the standard to verify the model´s accuracy and to practically examine the user interface´s feasibility. For scenario-based APS technology, the SAMI can construct the pre-defined scenarios model based on actual applications resource requests. Basically, the APS can set the server without workload to standby mode for power saving. Hence, the APS can automatically execute the power distribution optimization from SAMI. When the supervisor discovers any load changes under virtual environment, the wake-up function of the APS would activate the server and then execute live migration of the virtual machines (VM) by utilizing automatic workload balancer. This can achieve a workload balancing for the server and provide a virtualization environment with guaranteed performance for the network or cloud platform. By analyzing the experimental results, the proposed Scenario-Based APS can achieve a approximately 21% reduction in power consumption. Also, the operational performance of the virtual machine can also be safeguarded under the efficient operation of such power-saving model.