面向能耗的云工作流调度优化

云工作流调度直接决定了整个云工作流系统的性能,已成为一个重要研究内容.针对当前缺乏有效的面向能耗的云工作流调度优化方法,研究建立了面向能耗的云工作流过程模型、资源模型,提出了基于负载的能耗计算方法和面向能耗的云工作流调度优化算法.提出的方法考虑了能耗因素,在进行任务优先级计算及任务选择时不仅考虑了文件在不同虚拟机间传输的速度差异、同时考虑了虚拟机与本地共享数据库间的文件传输等因素,并从云工作流任务分配、主机负载和主机功耗关系的角度,进行能耗的计算与调度优化,更符合实际情况、使用范围更广.数值案例和仿真实验表明了提出的方法是可行的和有效的.

Scheduling for improving the energy efficiency of cloud workflow execution

As for cloud workflow systems, the scheduling directly determines the performance of whole system. The investigation on cloud workflow scheduling strategies is becoming a significant issue. However, there are still lack of the efficient methods to improve the energy efficiency of cloud workflow execution. In this paper, a process model and a resource model are introduced firstly for energy-aware workflow scheduling in cloud computing environments. Then, a method of calculating the energy consumption of cloud workflow execution is proposed based on host load. Finally, a scheduling algorithm is developed for improving the energy efficiency of cloud workflow execution. The proposed approaches have the following characteristics: 1) both the different data transmission rates among virtual machines and the data transfer between virtual machines and local shared database are considered; 2) the energy consumption of cloud workflow execution is estimated and optimized from the relationships among task assignment, host load and energy efficiency, which allows the proposed approaches for more wide scope of application. The numerical case and simulation experiments show the feasibility and effectiveness of the proposed approaches.