Improved PSO algorithm for energy saving research in the double layer management mode of the cloud platform
Ge Riletai, Jing Gao · 2016
The energy-saving research of virtualization of the cloud computing platform shows that there are problems in the management mode of the existing virtualization platform. This model is based on a single node managing the whole platform and the single model is responsible for migrating as well as scheduling all of the virtual machine. The migration scope of virtual machine for these models will be the whole platform, which leads to a great scale of migration problem, causes low efficiency and severely weakens the effect of energy saving. If the management node is low, the management function of platform will be failure. Therefore proposing a double management model of the virtual machine is used to solve the problem of single management node bottleneck and scope of the migration. At the same time, the improved PSO algorithm is used to make the plan for virtual machine migration. On the premise of meeting the service performance, the plan achieves energy saving by server booting to a minimum. Through the experiment, it proves that the proposed management mode not only solves the bottleneck problem of single management node, but also reduces the migration scope and the difficulty of the problem. The improved PSO algorithm obviously raises the speed of the migration and overall energy efficiency of scheme.