期刊文献+

HybSim:一种混合型流媒体存储的仿真系统 被引量:1

HybSim:A Simulation System for the Hybrid Streaming Media Storage
下载PDF
导出
摘要 高性能和低功耗的特点使闪存逐渐赢得广泛关注,而最常见的闪存使用方式是和内存、磁盘一起构成混合型存储系统,对于流媒体这样对存储容量和访问性能要求都非常高的应用来说尤其如此。但是,目前还缺少准确、全面的混合型存储仿真工具来支持混合存储的研究工作。首先,在对基于闪存缓存的混合存储的体系结构和能耗等关键问题进行建模的基础上,设计并实现了一套混合型流媒体存储的仿真系统HybSim。相对于目前国际上广泛使用的存储仿真系统DiskSim及其闪存补丁,HybSim增加了各类典型的混合型存储系统的实现、能耗模块、文件级管理和访问支持、流媒体服务模式支持,以及对性能、服务质量、能耗、设备耗损程度的统计模块等。基于HybSim对混合型流媒体存储系统在性能和节能两个方面进行了详细的仿真实验,并对几种典型的混合型存储系统体系结构在性能、服务质量、节能、设备磨损等方面的优劣进行了详细分析和比较。 The advantages of high-performance and low power consumption make flash to get more and more attention.The most common way of utilizing flash is constructing hybrid storage system together with RAM and disks,especially for streaming media application which has strict requirements on both storage capacity and performance.However,accurate and full-featured hybrid storage simulation tools are missing now.Firstly,this paper modeled the architectures and energy consumption of the hybrid storage systems,and then designed and implemented a hybrid storage simulation system———HybSim.Compared with the widely used storage simulation system——DiskSim and its flash patch,HybSim newly increases some new modules,including the implement of various kinds of hybrid storage architectures,power consumption,file-level management and accessing,streaming media service mode,and the statistics of performance,QoS,energy consumption and device wearing.Some detailed simulations on both the performance and energy conservation were made based on HybSim,and then we compared some typical architectures of hybrid storage on performance,Quality of Service(QoS),energy saving and device wearing.
作者 柴云鹏 杨楠
出处 《计算机科学》 CSCD 北大核心 2012年第9期1-4,19,共5页 Computer Science
基金 中国人民大学科学研究基金项目(中央高校基本科研业务费专项资金)资助
关键词 闪存 混合存储 仿真 流媒体 节能 Flash Hybrid storage Simulation Streaming media Energy saving
  • 相关文献

参考文献13

  • 1中国互联网络信息中心.中国互联网发展状况统计报告[EB/OL].http://www.cnnic.net.cn/dtygg/dtgg/201107/W020l10719521725234632.pdf.2011-7.
  • 2Yarow J. Videos on Youtube grew 123 year over year, while Facebook grew 239[EB/OL]. http://www, strangelove. eom/blog/2010/06/videos-on youtube=grew 123-yearover-year- while facebook grew-239,2010 06.
  • 3Gal E, Toledo S. Algorithms and data structures for flash memo- ries [J]. ACM Computing Surveys,2005,37(2): 138-163.
  • 4Ganger G, Ganger G, Worthington B, et al. The DiskSim simula- tion environment(v4. 0) [EB/OL]. http://www, pdl. emu. edu/ DiskSim, 2009-09.
  • 5Agrawal N, Prabhakaran V, Wobber T, et al. Design tradeoffs for SSD performance[C]//Proc. USENIX 2008 Annual Techni- cal Conf. , 2008. Berkeley, CA, USA: USENIX Association, 2008 : 57-70.
  • 6Kim Y, Tauras B, Gupta A, et al. FlashSim: A Simulator for NAND Flash-based Solid-State Drives [C]//Proc. First Int' 1 Conf. on Advances in System Simulation, 2009. Porto: IEEE, 2009 : 125-131 Bitar R_ Denlavin hvhrid storage Pools with Sun flash teehnolo-.
  • 7Bitar R. Deploying hybrid storage pools with Sun flash technolo- gy and the Solaris ZFS file system[EB/OL], http://wikis, sun. corn/download/attachments/190326221/820-5881, pdf, 2011-10.
  • 8Kgil T, Roberts D, Mudge T. Improving NAND Flash Based Disk Cache[C]//Proc. of the 35th Annual Int' 1 Syrup. on Computer Architeeture(ISCA) 2008. Beijing,China, 2008:327-338.
  • 9Matthews J,Trika S, Hensgen D, et al. Intel Turbo Memory: Nonvolatile Disk Caches in the Storage Hierarchy of Mainstream Computer Systems [J]. ACM Trans. Storage, 2008,4(2) : 24.
  • 10Pinheiro E, Bianchini R. Energy Conservation Techniques for Disk Array-Based Servers[C]//Proc. 18th Int'l Conf. Super- computing. New York, USA: ACM, 2004 : 68-78.

二级参考文献11

共引文献18

同被引文献3

引证文献1

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部