Some Energy Consumption Optimization Algorithm In Cloud Computing Data Center

In recent years, the scale of the servers and its corollary equipments is rapidly growing under the vigorous development of cloud computing in the world. However, quick response is needed while users asking for service in data center. So there are great many servers coordinating operation in cloud data center, which will bring high energy consumption. For low resource utilization and high energy consumption exist in data center, how to save energy consumption and operating costs become a big problem. As a result, energy-efficiency is a crucial issue in data centers of Data Clouds. At present, the energy-efficiency policies in cloud data center mainly contain the hardware energy-efficiency policy and the software energy-efficiency policy. The hardware energy-efficiency policy has reached a bottleneck in some way. However, load balancing can be optimized by the software energy-efficiency policy through virtualization technology, so the software energy-efficiency policy is considered as the most important energy-efficiency policy. VM consolidation, one of the software energy-efficiency policies, is considered as an effective policy to save energy. Thus, VM consolidation can be define as follows: on the premise of guaranteeing the quality of service(QoS), virtual machines of servers in data center are consolidated to a minimum number of servers by virtual machine migration, and then sleep/off the idle servers, which can make data center load balancing and save energy consumption by improving the utilization of resource in data center. VM migration mainly contains source host selection, VM selection and VM allocation. Some source host selection algorithms have been proposed CPU processing perspective. However, characteristics of jobs in data centers are neglected. In this paper, CPU utilization and RAM utilization are both considered to select the source hosts. Then, heuristic guided artificial bee colony algorithm is proposed to solve VM selection issue and VM allocation issue. With the idea of artificial bee colony algorithm, we respectively propose a VM selection algorithm(named as ABCS) and a VM allocation algorithm(named as ABCA). As a result, we propose a VM consolidation algorithm(named as ABCC) which combines the host selection algorithm, ABCS and ABCA. Compared with IQR_MMT, IQR_RS, IQR_MU, THR_MMT, THR_RS, THR_MU, MAD_MMT, MAD_RS and MAD_MU Consolidation algorithms in CloudSim3.0, the result shows that: on the promise of guaranteeing the quality of service, the proposed VM Consolidation algorithm, ABCA_ABCS, simulates that energy is saved 25%-30% and VM migration frequency is less than 1%.