Research On Virtual Machine Placement Strategy For Reducing Energy Consumption And Network Traffic Of Data Center

Cloud computing provides practical IT services to customers around the world,and it has a wide range of applications in the field of consumer,technology and business.With the rapid increase in the demand for cloud computing,the scale of the data center,which is managed cloud computing services,is growing sharply.The issue of the scalability in data center becomes prominent.And a large number of power consumption,resulting in high operating costs and carbon dioxide emissions.Therefore,the research on virtual machine(VM)placement in data center considering its energy efficiency and how to reduce the traffic in the network to enhance the scalability of data center can not only minimize operating costs,but also reduce the impact on the environment.At present,there are great limitations on the study of scalable and green data center.For example,when VMs are placed in data center,it is only considered the constraints of the server side energy,the network side energy or its scalability by the convergence of the network traffic.However,few study of the issue,the joint consideration of energy consumption of entire IT equipment and scalability of data center,has been proposed.In this thesis,the problem of green and scalability in virtualized data center is formulated.The energy efficiency of the data center and its scalable performance can be improved by integrating the resources of the server and reducing the data center network traffic when the VMs are placed in data center.It also points out that the virtual link mapping strategy between virtual machines can also affect the energy consumption of the network.The number of worked servers and network devices can be reduced by placing VMs properly and link mappings with appropriate policies and the amount of traffic in the network should be optimized in order to lower the network cost of the data center.The work of this thesis is as follows:1.Considering these optimization goals for data center servers,network energy consumption,and network costs,the author defines the problem of green and scalability in the virtualized data center and demonstrates that the problem is NP-complete.2.By analyzing the shortcomings of the existing data center energy saving algorithm,this paper proposes combined energy saving algorithm of VM placement based on energy consumption and scalability,and the greedy merge algorithm of link mapping.When the VMs are placed in the data center,the energy consumption of the server and the network can be reduced,and the traffic flow can be consolidated because the network traffic distribution will also affect the data center network energy efficiency and scalability.Simulation results show that the algorithm is effective.3.Different topologies are adopted in data center,which affects the location of the servers in the network,and then the traffic distribution in the network.So energy saving algorithm of VM tiering placement is proposed,which fully considers the structure of the Fattree topology and can further optimizes the traffic flow in the network.Comparing the algorithm proposed before,the network cost has been greatly improved,and the network energy efficiency has also been promoted.