混合架构下多请求模式的缓存替换模型研究

发布时间：2018-06-05 00:20

本文选题：访问模式 + 缓存策略　；参考：《计算机科学》2015年06期

【摘要】：针对多类型多访问模式应用的需求,在GDSF算法的基础上,引入平均访问间隔和最近访问间隔两个特性以增强算法的适应性;建立缓存结构模型,通过双关键字索引机制,快速索引缓存对象,降低系统开销;对超过一定大小的文件采取后缀预取策略以增加缓存中数据对象的个数。在课题应用背景下,与传统算法的对比实验表明,该方法能够减少缓存的平均请求等待时间,提高对象命中率和字节命中率,增强了缓存替换算法对多类型多请求模式应用的适应性。
[Abstract]:In view of the needs of multi type and multi access mode applications, on the basis of the GDSF algorithm, the average access interval and the nearest access interval are introduced to enhance the adaptability of the algorithm. The caching structure model is established, the cache object is quickly indexed by the double keyword index mechanism, and the overhead of the system is reduced, and the suffix of the files over a certain size is suffixed. The prefetching strategy is used to increase the number of data objects in the cache. Under the background of the subject application, the comparison experiment with the traditional algorithm shows that this method can reduce the average request waiting time of the cache, improve the hit rate and the hit rate of the object, and enhance the adaptability of the cache replacement algorithm to the multi type and multi request pattern application.
【作者单位】：上海大学计算机学院;
【基金】：国家“863”计划基金资助项目(2009AA012201-CFA2009SHDX01) 国家自然科学基金(61001163)资助
【分类号】：TP333

【参考文献】