摘要
Linux作为一个多任务、分时、通用的开源操作系统,越来越广泛地应用于各种商业和企业的服务器。为了提高系统的性能,Linux采用预取技术将应用程序所需的数据提前加载到缓存中,减少应用程序的I/O等待时间。然而由于服务器系统负载的多样化,导致了预取算法遇到越来越多的挑战。该文主要从分析Linux-2.6.29-rc2内核源代码入手,对Linux预取算法的体系结构与内部机制进行了深入分析与研究,并提出了一些改进预取算法的方法,对于进一步提高Lin-ux系统服务器的性能以及Linux的推广与使用具有重要的意义。
Linux is a multitask, timesharing , general open source operating system which is being applied for many servers of husiness and enterprises more and more . In order to improve system＇s performance, Linux introduces prefetching technology which can load data what application will accesss in future into cache so that applications can reduce their waiting time for I/O. However, system＇s loads vary more and more, so that prefetching algorithms face to more and more chanllenges. Firstly make an analysis for the source code about prefetching algorithm in Linux -2.6.29 - re2, to analyse and understand the architectures and internal mechanics of prefetching algorithm , then put forward some improved methods which have much important significances on improving system＇ s performance and spreading system＇s ranges.
出处
《计算机技术与发展》
2009年第12期93-96,共4页
Computer Technology and Development
