摘要
网络气象服务系统NWS产生的CPU可用性预测数据精确度高、可靠有效,已广泛应用于网格环境中。NWS是基于Unix/Linux操作系统实现的,在Windows平台下实现同样性能的CPU可用性的预测系统是必要的。论文设计并实现了基于NWS模型的Windows平台CPU可用性预测系统CAPredictor。实验显示,CAPredictor能够在Windows平台下提供CPU可用性的精确预测数据,提高用户和调度进程选择CPU资源的效率。
The predictions of CPU availability which are generated by the network weather service system（NWS） are accurate,credible and effective to grid users.NWS is currently deployed for Unix/Linux system so it is necessary to develop a system for Windows which can generate the same useful CPU availability data.This paper introduces the designs of the CPU Availability Predictor for Windows which is on the basis of NWS＇s computing and forecasting models. The experiment results show that CAPredictor can make the accurate predictions of CPU availability and increase the efficiency of users＇and schedules＇CPU resources selection.
出处
《计算机工程与应用》
CSCD
北大核心
2006年第10期94-96,共3页
Computer Engineering and Applications
基金
国家自然科学基金资助项目(编号:60473099)
吉林省杰出青年基金(编号:20040119)
