期刊文献+

统一缓存:基于用户层通信的合作缓存技术 被引量:1

Cooperative Caching Based on User-Level Communication
下载PDF
导出
摘要 合作缓存机制是集群系统提高整体性能的一种有效方法 ,其利用高速网络将各个结点的缓存进行合作管理与访问 ,大幅度提高了缓存的命中率 但传统的合作缓存技术没有考虑到广为应用的高效用户层通信机制的特点 提出一种新的用户层通信与合作缓存技术相融合的缓存机制———集群统一缓存 这一机制充分利用了用户层通信的特点 ,包括协议精简、零拷贝、虚拟内存映射通信 (VMMC)技术等 ,将缓存与结点通信相融合 ,减少了集群系统应用程序IO模块的层次与复杂度 ,提高了系统性能 同时这一机制也顺应了IO子系统日益独立化的发展趋势 该技术已经应用于自行开发的面向对象的Internet服务存储平台———TODS上 ,具有高效。 Cooperative caching in cluster system is an efficient mechanism to improve the performance of the whole IO sub system, which uses high speed network to combine the local cache of each node into a coordinated global cache with higher hit rate However, the traditional cooperative caching does not take into account the features of user level communication employed more and more widely today Presented in this paper is a new cooperative caching mechanism-uniform caching, which employs user level communication technologies to unite both the communication layer and the caching module Based on some mechanisms including simplified protocol, zero copy, and virtual memory mapped communication, uniform caching predigests the IO sub system and increases the whole performance Moreover, this caching mechanism has been implemented on an object oriented cluster storage platform for Internet applications-TODS
出处 《计算机研究与发展》 EI CSCD 北大核心 2003年第7期1117-1123,共7页 Journal of Computer Research and Development
基金 国家自然科学基金 ( 60 2 0 3 0 2 3 )
关键词 用户层通信 合作缓存 虚拟内存映射通信 GM user level communication cooperative caching virtual memory mapped communication GM
  • 相关文献

参考文献11

  • 1M D Dahlin, R Y Wang, T E Anderson, D A Patterson.Cooperative caching: Using remote client memory to improve file system performance. In: Proc of the 1st Symposium on Operating Systems Design and Implementation. San Diego: USENIX Press,1994. 267--280.
  • 2E A Thomas, E C David, D A Patterson. A case for networks of workstations: NOW. IEEE Micro. 1995, 15(1): 24--31.
  • 3T Anderson, M Dahlin, N Jeanna et al. Serverless network file systems. In: Proc of 15th Symposium on Operating Systems Principles. Philadelphia: MIT Press, 1995. 21--35.
  • 4P Sarkar, J Hartman. Efficient cooperative Caching using hints.The USENIX' 1996, Seattle, Washington, 1996.
  • 5S Araki, A Bilas, C Dubnicki et al. User-space communication: A quantitative study. The Supercomputing 98 Conf (SC'98).Orlando, Florida, 1998.
  • 6Doug Johnson. GM: The low-level message-passingsystem for myrinet networks. 2001. http://www, myricom, com/scs/index.html.
  • 7Craig Russell. DAFS Specifications. 2001. http://www.dafscollaborative.org/tools/spec, shtml.
  • 8Robert Grimm. Myrinet performance measurements. 2001.http://www, myricom, com/myrinet/perfor-mance/index, html.
  • 9S H Rodrigues, T E Anderson, D E Culler. High-performance local-area communication using fast sockets. The 1997 USENIX Cord, Anaheim, C A, 1997.
  • 10C Dubnicki, A Bilas, Y Q Chen et al. VMMC-2: Efficient support for reliable, connection-oriented communications. Hot Interconnects'97, Stanford, CA, 1997.

同被引文献3

引证文献1

二级引证文献4

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