基于云模型的网格QoS参数归约与调度优化

Grid resource schedule optimizing based on QoS parameters induced by cloud model

针对网格资源调度中用户对QoS的定性描述,利用云模型实现资源调度中的QoS匹配。深入分析了QoS参数云的特征,提出了QoS云处理模型,通过该模型,将离散的多个QoS参数归约到一个定性的概念上;设计了实现参数归约的体系结构;给出了基于定性概念的资源调度算法。实验表明,所提出方法在资源调度率和吞吐量以及系统资源的利用效率等方面体现出良好的特性,实现了基于定性概念的调度,达到了优化调度的目的。

QoS parameters is matched one by one in current grid resource scheduling,the trouble of this task only offers a qualitative concept about QoS requirement.In this paper,a new approach is presented which empowered QoS parameters is matched by qualitative concept that tasks offer.Firstly,QoS parameters based on the cloud model are analyzed,and induce model of QoS is proposed,by this,QoS parameters can be induced to a qualitative concept.Secondly,the architecture of QoS transform is presented.At last,the optimize scheduling algorithm is described based on the model.Finally the simulations are performed to compare the performance of optimize scheduling algorithm with that of the time minimisation and cost minimisation.The experiments show that the optimize scheduling algorithm is efficient on throughput and system resource utilization ratio within grid computing environments.