数据中心电费最小化的负载调度策略

Workload Scheduling for Minimizing Electricity Cost of Data Center

为了降低电费成本,一些数据中心使用绿色能源供电。然而,负载的波动性和电价的时间差异性给数据中心电费成本带来了挑战。针对上述问题,提出一种低成本的负载调度算法,使得数据中心电费最小化。首先,建立电力耗费模型;然后,将电费最小化问题形式化为一个多目标约束的优化问题;最后,求解该优化问题得到相应的负载调度策略。实验结果表明:该算法可以在保证负载性能的前提下,有效降低数据中心的电费成本。

In order to reduce both electricity cost and carbon emission,some data centers begin to use green energy supply. However, challenges arise with the fluctuating workload and temporally diverse electricity price to data center electricity cost. To deal with these challenges, this paper presented a workload scheduling algorithm which can minimize the total electricity cost was data center. First,a model for the total electricity cost was introduced. Then a multibjective constrained optimization problem to the electricity cost was formulated. Finally, the corresponding workload scheduling policy was formed by solution of the optimation problem. Experimental results show that the proposed algorithm can effectively reduce the total electricity cost.