一个面向数据密集型应用程序的网格工作流调度方案

A Case of Workflow Scheduling Method in Grid for Data Intensive Applications

采用网格工作流调度和数据感知调度相结合的方法,提出一个面向数据密集型应用的网格工作流调度方案.在CSF4网格元调度程序中以调度插件的形式实现了该方案,并且保证两个调度插件之间能够协同工作.实验结果表明,该方案能有效解决数据密集型网格工作流作业的调度问题,并且进一步完善了CSF4多调度插件协同工作机制.

The authors proposed a case of workflow scheduling method in grid for data intensive application via both workflow scheduling and data-aware scheduling. We implemented this solution by means of scheduling plugins in meta-scheduler CSF4, ensuring the two scheduling plugins work together coordinately. The experimental results show that this method solves the scheduling problem about data-intensive workflow applications efficiently and also improves the scheduling multi-plugin mechanism of CSF4.