一种基于资源状态预测的资源调度模型

Resource scheduling model based on resource state prediction

网格资源管理系统在进行任务请求与资源的匹配调度时,应考虑资源的网格环境以及资源的当前工作状态的问题.提出了一种基于资源状态预测的资源调度模型,结合资源状态和距离计算资源的共享开销,并作为任务资源匹配调度的依据,从而为任务请求分配“最好”的共享资源.

When making matching and scheduling between grid resources and requests, the resource management system should take the environment and running state of resources into account. A resource matching and scheduling model based on resource state prediction is proposed. By taking the resource distance and running state into share cost calculation, the resource scheduler can find the ＂best＂ resource for user＇s requests.