A Resource Scheduling Strategy for the CFD Application on the Grid

In this paper, we focus on the scheduling issue for one kind of high performance computing applications, that is, computational fluid dynamics applications. Firstly, we focus on studying the characteristic of the CFD applications, and model this kind of applications that can be decomposed into multiple sub-jobs (tasks), which evolve to be represented by a DAG. Then, a hierarchical infrastructure for resource organization in the computational grid environment is proposed. Thirdly, we discuss the scheduling strategy in the presented scenario, and propose a task scheduling framework and analyze a corresponding algorithm with simulation experiments. Finally, we discuss the implementation with the related project.