网格计算中一种改进的工作流调度算法

An Improved Workflow Scheduling Algorithm in Grid Computing

网格计算中如何有效地实现工作流的调度问题是目前的研究热点。文中综合考虑了资源节点的动态负载和服务能力,提出了一种改进的调度算法(AWSA)。该算法首先对任务的优先级进行降序排列,然后依次为它们选择具有最大截止时间约束的服务站点作为其候选资源,最后,依据资源站点的任务分配情况和负载变化趋势,自适应地实现从任务资源请求到站点的映射。仿真实验结果表明,文中方法是有效的,在作业拒绝率和作业调度长度方面,AWSA的性能要优于已有的方法。

How to effectively implement the workflow scheduling problems in a grid computing is currently a hot topic. Take into the dynamic load and service capabilities of the resource nodes account, propose an improved workflow scheduling algorithm （A WSA） . Firstly, the priority of the task is descended, and then followed them to choose the service site with the maximum cut-off time constraints as their candidate resources, finally, according to the task allocation and load trends of the resource site, ASW A can adaptively achieve the mapping from the resource requests of task to the service site. The experimental results show that this method is effective, and the job rejection rate and job schedule length of A WSA is superior to the existing methods.