Energy-consumption optimization method for cloud computing platform operation considering random tasks

摘要

Energy efficient problem is one of the main obstacles to open the new era of green-cloud calculation. Especially with the global climate change and the increase of dependence on energy, green computing receives extensive attention as a new high-end calculation. In order to realize the minimization of energy consumption of cloud computing, for the problems that cloud computing system generates lots of free energy consumption in operation process because of idle compute nodes, and the unmatched task scheduling generates waster of energy optimization consumption, the paper proposes an energy optimization management method based on power control of server of CPU utilization ratio and task scheduling strategy. Firstly, the queuing model is used to model cloud computing system. The average response time and average power of cloud computing system is analyzed, and the energy consumption model of cloud computing system is established. Then, based on CPU utilization ratio, task scheduling strategy with large-service intensity and small -execution energy consumption, the idle energy consumption and luxury energy consumption receives optimization control. Based on the scheduling strategy, the paper designs BC-MC Algorithm. With the minimal expected execution energy consumption meeting performance constraint. The experimental results indicate that the algorithm can reduce the expenditure of energy consumption of cloud computing system under the premise execution performance.