摘要
如何在动态网格环境中获取网格应用的更好性能是网格技术一直关注的课题 为实时地调整和改善应用性能 ,介绍了一种基于性能监控的网格应用自适应调节机制 ,该机制运用模糊控制技术 ,不依赖于应用对象的数学模型 ,对无法建模或很难建模的复杂对象 ,能利用人的经验知识完成调节任务 结合网格监控体系结构 (GMA) ,设计了一个实现该机制的模型 ,描述了模型中各模块的具体功能 ,给出了实现技术及相关算法 ,最后对这种自适应调节机制的性能进行了评估 该模型结构建立在GMA基础之上 。
How to obtain better performance of grid application in the dynamic grid environment is always a focus of grid technology.For real-time tuning and improving application performance, an adaptive tuning mechanism for grid application based on performance monitoring is introduced.This mechanism adopts the fuzzy control technology, and can complete some tuning tasks on difficult-to-make models using human experience and knowledge without depending on math model of the grid applications.A model of adaptive tuning mechanism is designed based on grid monitoring architecture (GMA), the function of each module in the model is depicted, and the technology and algorithm of model realization is described.Finally, an experiment is conducted for evaluating the performance of the mechanism and the experimental result is presented via a visualization tool.Because the model of adaptive tuning is based on GMA, it is provided with better scalability.
出处
《计算机研究与发展》
EI
CSCD
北大核心
2004年第12期2175-2180,共6页
Journal of Computer Research and Development
基金
国家"八六三"高技术研究发展计划引导项目(2003AA001032)
软件工程国家重点实验室开放基金项目(SKLSE0314)
关键词
网格
自适应调节
模糊控制
性能监控
grid
adaptive tuning
fuzzy control
performance monitoring
