Study On Energy Consumption Optimization For Cloud Data Center

Cloud computing as a business computing model and service model,has drawn much attention in the present moment,and in recent years,with the increase of the number of the cloud data center,the problem of high energy consumption is particularly prominent.Based on the data center energy consumption optimization as the research background,in order to reduce the data center PUE value for research purposes,to virtual machine migration and online resource scheduling two energy saving strategy for optimizing method,and made the following several aspects work:1.According to the characteristics of cloud data center two energy estimation model is put forward that the energy consumption model based on the server and to estimate the energy consumption model based on different states.2.The energy consumption model based on the server,energy-saving scheduling algorithm of this paper is based on the virtual machine migration for the backing,through absolute position difference to determine whether the server overload;Finally,through the perception of the target server selection algorithm of energy consumption,the load is lower than the lower limit on the threshold of the server to migrate virtual machine and shut down the server.3.Based on the energy consumption of different state estimation model,this article is based on resources in different conditions such as sleep,wake up,work cost is different,and by measuring the socket for estimating validation and under different resource utilization by the same task of testing,verify the relationship between resource utilization and energy consumption,power consumption estimation model is established4.According to the energy consumption of the server,in this paper,the improved perception algorithm and resource scheduling algorithm,energy consumption of the two types of estimation model is set up in the center of the cloud computing in CloudSim simulation experiments on the simulation platform.By experimental verification,in this paper,the improved energy consumption perception algorithm is effectively reduced the number of virtual machine migration.At the same time,the improved Min-Min resource scheduling algorithm,by comparing the minimum total power consumption needed to complete the task scheduling,the scheduling algorithm of the test,to verify the algorithm on the energy-saving effect.