一种基于协作型任务的网格资源调度算法

网格是建立在互联网上的新一代基础设施,网格计算中大部分任务为子任务之间具有复杂关系的协作型任务。首先对协作型进行了详细的分析,采用AOV网表示子任务之间的关系,针对协作型任务提出了一种基于遗传模拟退火算法的调度策略,该调度策略满足了对协作型任务资源调度的需求。同时还阐述了该算法的原理和操作步骤。

Scheduling in Grid Computer Environment Based on Collaborative Task

Grid is the new generation of infrastructure established on the Internet,in the grid computing the majority of tasks are the collaborative tasks composed by some sub-tasks with complex relations. In this paper we make a detailed analysis of such kind of tasks and adopt the AOV net to express the relations between the sub-tasks. We propose a scheduling strategy based on the genetic simulation annealing for the collaborative task, and elaborate the principle and the steps of this strategy. The scheduling satisfies the demand of scheduling for the collaborative task.