异构云环境下基于分簇的云资源感知任务调度方案

Task scheduling scheme based on clustering in heterogeneous cloud computing platform

针对提高异构云平台中资源调度的效率,提出了一种基于任务和资源分簇的异构云计算平台任务调度方案。利用K-means算法,根据任务的CPU和I/O处理时间对任务分簇,根据资源的计算能力对资源分簇;然后,将任务簇对应到合适的资源簇,并利用最早截止时间优先(EDF)算法对任务簇中的独立任务进行调度,利用提出的改进型最小关键路径(MCP)算法对依赖性任务进行调度。实验结果表明,在资源异构的云计算环境中,该方案执行任务时间短、能耗低。

To improve the efficiency of resource scheduling in heterogeneous cloud platform, a task scheduling scheme based on task and resource clustering in heterogeneous cloud computing platform was proposed. First, clustered the tasks according to the CPU and I/0 processing time of tasks with the K-means, and clustered the resources according to the computing power of resources; then, it made the task cluster corresponding to the appropriate resource cluster, and the earliest deadline first （EDF） algorithm was used for the independent task scheduling, and the improved minimal critical path （MCP） algorithm was used for dependent task scheduling. Experimental results show that in the cloud computing environment with heterogeneous re- sources the proposed scheme takes more short time and consume lower energy during the task execution.