论文部分内容阅读
针对网格环境下计算节点的自治性、异构性、分布性等特征,提出了一种动态的基于任务响应时间预测的调度算法。该调度方法依据历史数据和最近访问过计算节点的任务请求提交时间、任务完成时间、网络通信延迟等信息,预测计算节点将来的任务响应时间,将任务提交给轻负载或性能较优的计算节点完成。实验结果表明,该方法不但可以有效减少不必要的延迟,而且在任务响应时间、任务的吞吐率及任务在调度器内等待被调度的时间方面比随机调度等传统算法要优。
Aiming at the autonomy, heterogeneity and distribution of computing nodes in grid environment, this paper proposes a dynamic scheduling algorithm based on task response time prediction. The scheduling method predicts the future task response time of the compute node based on the historical data and the request time, task completion time and network communication delay of the last visited compute node, and submits the task to the light load or better-performance compute node carry out. Experimental results show that this method can not only reduce unnecessary delay effectively, but also outperform traditional algorithms such as stochastic scheduling in terms of task response time, task throughput rate and task waiting time in scheduler.