Research On Virtual Resource Energy Efficiency Model And Scheduling Algorithm Based On QoS And Energy Consumption Weighing

Cloud data center as an important part of cloud computing,has replaced the traditional data center.Because of building and operation of the large scale cloud data center,there is too much energy wasted and carbon dioxide emissions.Especially the lower resource utilization,it results in the environmental pollution and a waste of resources.The rapid development of virtual technology has provided an effective solution for that.A lot of researchers proposed many feasible solutions.After comparing the advantage of different methods,this thesis proposes the EEQoS model(The energy-efficient QoS)that could balance energy consumption and QoS of user.Moreover,using the strategy based on migration overhead minimum optimizes the particle swarm migration algorithm using minimum migration strategy.Our contributions in this thesis are:(1)Constructing EEQoS Model with Weighing Energy and QoS.High energy consumption research in improving the efficiency of virtual resources often pay more attention to performance indicators,and then the relevant research shows that performance can not accurately describe the status of QoS.In addition,there is no standard user QoS index in the field of cloud computing.Therefore,this thesisrefers to the results of other studies,selects the parameters such as delay,cost and throughput,and quantifies the parameters,and evaluates the QoS satisfaction of users.At the same time,the energy consumption based on multi-resource is added,and the user satisfaction of unit energy consumption is taken as the basis for evaluating the energy consumption of cloud data center.(2)Design the smallest MCost strategy based on migration overhead.Dynamic migration of virtual machines is an effective means of energy efficiency management,but unreasonable virtual machine migration algorithm will affect the performance of cloud data center,thus affecting the user's QoS.The most important step in the virtual machine migration algorithm is to select the virtual machines that need to be moved out of the physical nodes.Most of the previous studies are based on the FFD algorithm,but the strategy may cause a large migration overhead.Virtual machine migration costs,select the need to move out of the virtual machine.(3)Verify the EEQoS model and the MCost policy.This thesis validates the advantages and disadvantages of the EEQoS model by checking the cost,response time and throughput.The performance of the virtual machine migration algorithm based on the migration cost is verified by the energy consumption,the number of physical nodes enabled and the number of virtual machine migration.The experimental results show that the EEQoS model and the MCost strategy are better than the ECED model and the FFD algorithm based strategy.