云环境下数据密集型工作流调度

Data intensive workflow scheduling in cloud environment

相比于传统的计算密集型工作流,云环境下数据密集型工作流中数据传输量大,任务数量多,任务之间依赖关系复杂。为此,文章提出了一种基于最大数据传输路径的数据密集型工作流调度方法,首先对工作流进行分层,找出工作流中最大数据传输量路径,并开展对最大数据传输量路径的分配工作,然后对其余任务分配服务商,以此缩短工作流的完成时间,最后通过具体实验验证了该方法的可行性和有效性。

Compared with the traditional computing intensive workflow,the data intensive workflow in cloud environment has a large amount of data transmission,a large number of tasks,and complex dependencies between tasks.Therefore,this paper proposes a data intensive workflow scheduling method based on the maximum data transmission path.First,the workflow is layered to find the path of the maximum data transmission in the workflow,and the allocation of the path of the maximum data transmission is carried out first,Then,we assign service providers to other tasks to reduce the completion time of the workflow.Finally,we demonstrate the feasibility and effectiveness of this method through specific experiments.