摘要
根据存储系统各级缓存的不同特点,提出了一种基于网络的异构协同预取方案RLCP(RA and Linux collaborate prefetching),它通过在存储客户端和存储服务器端使用不同的预取算法和替换算法,构建一个异构的两级存储系统,并利用合适的协同预取策略,动态调整预取算法的积极性,提高了存储系统的性能,同时也为云存储提供了一种优秀的解决方案。
According to the different characteristic of the cache in each level, akindofRLCP （RAandLinuxcollaborateprefetching） is proposed based on network. Different replacement algorithm and prefetching algorithm is used to architect a heterogeneous two-level storage system. Through appropriate collaborate prefetching strategy which dynamically adjusts the aggressiveness ofthe server-side prefetching activities, the performance of the storage system is enhanced, and it is also an excellent scheme for cloud storage.
出处
《计算机工程与设计》
CSCD
北大核心
2010年第2期341-343,共3页
Computer Engineering and Design
基金
北京市优秀人才基金项目(Q0007013200801)
国家自然科学基金项目(60873145)
关键词
多级存储系统
网络存储
云存储
存储优化
异构协同预取
multi-level storage system
network storage
cloud storage
storage optimization
heterogeneous collaborate prefetching
