A novel task deployment approach based on graph theory for power saving
Gaochao Xu1, Peng Liu1, Xiaodong Fu1, Yunmeng Dong1, Jia Zhao2, Yan Ding1
COMPUTER MODELLING & NEW TECHNOLOGIES 2014 18(5) 73-78
1 College of Computer Science and Technology, Jilin University, Qianjin Str. 2699, 130012 Changchun, China
2 College of Computer Science and Engineering, ChangChun University of Technology, Yan’an Str. 2055, 130012 Changchun, China
With the increasing of the big datacenter, the power consumption seems to be another overhead except the equipment cost. Saving the power of big datacenter is the hotspot now. In this paper, we proposed TA-BG algorithm based on the linear weighted and graph theory to speed up the execution of tasks. Firstly, utilizing linear weighted to execute first filter to reduce the searching scope for the next research. Secondly, seeking out the hosts that can execute tasks fast based on graph theory. Finally, placing the host on the hosts selected above. The experiments indicate that TA-BG can save power of datacenter by reducing the executing time. Besides, the TA-BG even performs well on load balance.