Workflow Task Scheduling Strategy In Cloud Computing Environment

Presently, with the continuous development of the cloud computing, a series oftechnical problems need to solve urgently in the development of cloud computing tomeet the demands of its development. A typical problem is how to reduce the costsand achieve revenue maximization under the premise of maintaining qualifiedservices. In the process of solving this problem, the concept of the workflow oncloud computing platform has been raised. On the one hand, such workflow providesthe abstract definitions of tasks, the flexible configuration and the automaticoperation to improve the quality of services; on the other hand, it facilitates the taskmanagement and resource optimization in the scheduling process and hence reducesthe cost. Thus the workflow on the cloud computing platform, as an importantsolution to the issues of the cloud computing, has drawn more and more people’sattention to study it. In the study of the workflow, task scheduling will directlydetermine the success rate and efficiency of its execution. Therefore, in thisdissertation, the task scheduling in cloud computing environment carried on anin-depth research.In this dissertation, combined with the characteristics of cloud computing whileaiming at the phenomena of idle resources and low-reliability workflow emerged inthe scheduling process, the two-step task scheduling strategy for the workflow onthe cloud computing platform and the reliability-aware workflow scheduling strategyon the cloud computing platform are put forward. The former one, fully combinedwith tasks relationships in the workflow, adopted the idea of inserting tasks in theidle time of resources and assigning tasks to appropriate resources to achieve theshorter completion time and better utilization of resources. The latter one was toalways consider the reliability of the workflow. Meanwhile it enhanced thetransmission reliability of the task by the copying strategy and reduced thetransmission time between the tasks to achieve the shorter completion time and higher workflow reliability.The simulation experiments showed that compared with the existing algorithmsthe two-step task scheduling strategy for the workflow on the cloud computingplatform had achieved the shorter completion time and better utilization of resources;and the reliability-aware workflow scheduling strategy had achieved the shortercompletion time and the higher workflow reliability.