期刊文献+

TCP Offload系统的设计与实现 被引量:1

Design and Implementation of TCP Offload System
下载PDF
导出
摘要 网络带宽的迅速发展使主机协议处理开销成为系统整体性能的瓶颈。为了提高网络应用的性能、降低主机系统的资源占用,该文采用网卡子系统分担主机的网络包收发和TCP协议处理,设计实现了一个TCP offload系统,分析了一般TCP offload系统在性能和实施的困难,给出系统设计的基本原则,介绍了系统在通信接口、资源管理、事件管理和用户接口等方面的实现。系统测试结果表明,网卡能处理高速的网络应用,在一些情况下甚至达到线速。 With the increase of network bandwidth, protocol overheads on host have become the bottleneck of system performance. To improve the performance of network applications and reduce the consumption of host resources, this paper exploits the network subsystem to offioad the receiving and transmitting of packets and the processing of transmission protocol stack. It implements the TCP offioad system and analyses the difficulties of general offioad system in performance and deployment. It describes the basic rules of system design, the implementation of communication interface, resource management, event management and application interface. Tests show that the NIC can process high-speed network application, and in some cases it can handle line-speed traffic.
出处 《计算机工程》 CAS CSCD 北大核心 2007年第24期99-101,共3页 Computer Engineering
基金 国家"863"计划基金资助项目"面向网络的高性能计算机--曙光4000"(2002AA104410)
关键词 TCP下移 高速网络处理 协议栈 TCP offload high-speed network processing protocol stack
  • 相关文献

参考文献5

二级参考文献10

  • 1Scalable Coherent Interface[S].IEEE Std 1596--1992,IEEE CS Press, 1993.
  • 2N J Boden et al.Myrinet:A Gigabit Per-Second Local-Area Network [J].IEEE Micro, 1995 ; 15( 1 ) :29-36.
  • 3Fabrizio Petrini,Wuchun Feng et al.The Quadrics Network:High- performance Clustering Technology[J].IEEE Micro,2002:46-57.
  • 4Intel Corporation.i960(r)VH Processor Developer's Manual.1998-10.
  • 5Sumimoto S, Tezuka H. The Design and Evaluation of High Performance Communication Using a Gigabit Ethernet. ICS '99 Rhodes, Greece, 1999:260-267
  • 6Schroeder P J, Korel B.Black-box Test Reduction Using Input-output Analysis. ISSTA, Portland, Oregon,2000: 173-177
  • 7Santhanam U. Automating Software Module Testing for FAA Certification. Proc. SIGAda, Bloomington, MN, USA, 2001:31-37
  • 8Buck B R, Hollingsworth J K. Using Hardware Performance Monitors to Isolate Memory Bottlenecks. Proceedings of Supercomputing 2000, DALLAS, Texas, USA , 2000
  • 9O'Carroll F, Tezuka H. The Design and Implementation of Zero Copy MPI Using Commodity Hardware with a High Performance Network. ICS98 Melbourne Australia, ACM, 1998:243-250
  • 10安学军,祝明发,高文学,吴冬冬.虫洞路由交换及其缓冲区设计[J].微电子学与计算机,2002,19(9):15-18. 被引量:4

共引文献5

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