TCP性能的应用透明优化
|
摘要
当前的应用越来越呈现分布式的特点,这些应用主要是由IP网络进行承载。由于IP网络尽力而为的特性,决定了其总是表现出一定的损伤,如延迟、抖动和丢包等,这些网络损伤会极大地影响分布式应用的性能。
论文的研究目的是希望通过对TCP性能优化的应用透明实现,提高分布式应用对网络损伤的适应能力,从而改善它们的性能表现。论文的研究内容包括:分析TCP协议以及其拥塞控制机制,找出从协议层面进行控制存在的不足以及应用透明实现TCP性能优化的必要性和可能性,给出一些实现优化的具体方法,包括滑动窗口大小的修改、选择确认、数据包压缩、数据包合并和数据包层面的前向纠错等,此外,论文还讨论了Windows平台下数据包的处理方法。
论文的主要成果包括:以Windows2003 Server作为实现平台,通过网络驱动接口规范(NDIS)实现了对数据包进行截获、修改和发送,在此基础上实现了TCP性能的应用透明优化的各种具体方法,并进行了测试以验证优化结果,最后结合测试结果和最新的技术发展,讨论了优化工作仍然存在的不足,并指出了下一步的改进方向,包括处理能力的提升、数据识别基础上的流量缩减、QoS、数据压缩的改进和对UDP的支持等。
论文系统讨论了在TCP性能优化应用透明的实现方式和方法,以及在Windows下进行数据包的截获、修改、发送等技术,这些具有一般性的借鉴意义。测试结果表明,论文的研究成果在克服网络延迟、丢包等方面达到了预期的效果,可以较好地提高分布式应用吞吐量和反应时间等指标。
The applications nowadays are more and more distributed and are mostly carried by IP network. The IP network always has some impairment, such as delay, jitter and packet loss, due to its best effort characteristic. Those network impairments will affect the performance of those distributed applications badly.
The aim of this paper is to improve the capability of distributed applications when faced with network impairments by the application-transparent realization of TCP performance optimization and then hence to improve the performance of those applications. The paper studies TCP and its congestion control mechanism, the shortages TCP congestion control mechanism at protocol level still has nowdays, the necessity and possibility of application-transparent realization of TCP performance optimization. Then the paper discusses the techniques to achieve such optimization, including congestion window re-sizing, selective acknowledgement, data compression, packets coalescing and FEC (Forward Error Correction) at packets level. Besides, the paper also discusses how to deal with packets in Windows.
The paper achieves packets capture, modification and forwarding by NDIS in Windows2003 Server environment. Based on the above, the paper accomplishes the application-transparent TCP performance optimization techniques and verfies the optimization result by testing. The paper further discusses the deficiency of the optimization work has been done, based on the testing results and the latest technology, and then points out those aspects to be improved in the next stage, which include processing capability improvement, data reduction based on data recognition, QoS, improvement at data compression and UDP support.
The paper has comprehensively discussed the way and the mothods to achieve application-transparant TCP performance optimization and also how to capture, modify and transfer packets in Windows environment. These are referential. The testing results show that the work of the paper has achieved what has been expected at network delay and packet loss mitigation and can improve the application and response time remarkably.
论文的研究目的是希望通过对TCP性能优化的应用透明实现,提高分布式应用对网络损伤的适应能力,从而改善它们的性能表现。论文的研究内容包括:分析TCP协议以及其拥塞控制机制,找出从协议层面进行控制存在的不足以及应用透明实现TCP性能优化的必要性和可能性,给出一些实现优化的具体方法,包括滑动窗口大小的修改、选择确认、数据包压缩、数据包合并和数据包层面的前向纠错等,此外,论文还讨论了Windows平台下数据包的处理方法。
论文的主要成果包括:以Windows2003 Server作为实现平台,通过网络驱动接口规范(NDIS)实现了对数据包进行截获、修改和发送,在此基础上实现了TCP性能的应用透明优化的各种具体方法,并进行了测试以验证优化结果,最后结合测试结果和最新的技术发展,讨论了优化工作仍然存在的不足,并指出了下一步的改进方向,包括处理能力的提升、数据识别基础上的流量缩减、QoS、数据压缩的改进和对UDP的支持等。
论文系统讨论了在TCP性能优化应用透明的实现方式和方法,以及在Windows下进行数据包的截获、修改、发送等技术,这些具有一般性的借鉴意义。测试结果表明,论文的研究成果在克服网络延迟、丢包等方面达到了预期的效果,可以较好地提高分布式应用吞吐量和反应时间等指标。
The applications nowadays are more and more distributed and are mostly carried by IP network. The IP network always has some impairment, such as delay, jitter and packet loss, due to its best effort characteristic. Those network impairments will affect the performance of those distributed applications badly.
The aim of this paper is to improve the capability of distributed applications when faced with network impairments by the application-transparent realization of TCP performance optimization and then hence to improve the performance of those applications. The paper studies TCP and its congestion control mechanism, the shortages TCP congestion control mechanism at protocol level still has nowdays, the necessity and possibility of application-transparent realization of TCP performance optimization. Then the paper discusses the techniques to achieve such optimization, including congestion window re-sizing, selective acknowledgement, data compression, packets coalescing and FEC (Forward Error Correction) at packets level. Besides, the paper also discusses how to deal with packets in Windows.
The paper achieves packets capture, modification and forwarding by NDIS in Windows2003 Server environment. Based on the above, the paper accomplishes the application-transparent TCP performance optimization techniques and verfies the optimization result by testing. The paper further discusses the deficiency of the optimization work has been done, based on the testing results and the latest technology, and then points out those aspects to be improved in the next stage, which include processing capability improvement, data reduction based on data recognition, QoS, improvement at data compression and UDP support.
The paper has comprehensively discussed the way and the mothods to achieve application-transparant TCP performance optimization and also how to capture, modify and transfer packets in Windows environment. These are referential. The testing results show that the work of the paper has achieved what has been expected at network delay and packet loss mitigation and can improve the application and response time remarkably.
引文
[1] 陈向群, 杨芙清. 操作系统教程. 北京:北京大学出版社,2004.
[2] InfoWorld.Solving the Branch Office Dilemma:Shifting from Tactical to Strategic Investment,2005-10.
[3] The Strategy Group.Avoiding the Branch Office Quagmire,2005-08.
[4] Cerf,V.,Kahn,R.A protocol for Packet Network Interconnection. IEEE Transactions on Communications. 1974.
[5] I.113. Vocabulary of Terms for Broadband Aspects of ISDN. CCITT. 1988.
[6] I.121. Broadband Aspect of ISDN. CCITT. 1988.
[7] R. Braden, D. Clark, S. Shenker. Integrated Service in the Internet Architecture: An overview. RFC 1633, 1994.
[8] L.Zhang, S. Berson, S. Herzog, S. Jamin. Resource ReSerVation Protocol (RSVP) -Version1 Functional Specification. RFC 2205, 1997.
[9] S. Blake,D. Black, M. Carlson,E. Davies,Z. Wang,W. Weiss. An Architecture for Differentiated Service. RFC 2475, 1998
[10] William Stallings. High-Speed Networks and Internet――Performance and Quanlity of Service, 齐望东,薛卫娟,傅麒麟,胡谷雨等译,谢希仁审校. 北京:电子工业出版社,2003. P239~240
[11] ITU-T. Recommendation Y.1541. Network performance objectives for IP-based services,2002-05:P7
[12] Ceorge Coulourisv,Jean Dollimore,Tim Kinderg. Distributed Systems Concepts and Design Third Edition, 金蓓弘等译. 北京:机械工业出版社,中信出版社,2004. 12~18.
[13] Mark Fabbi,Latency: The Silent Killer of Application Performance. 2004-9-14.
[14] Shunra Software Ltd,The Network Impact: Extending load testing to ensure application performance for remote end users. 2004.
[15] K Claffy,G. Miller,K. Thompson. The Nature of the Beast: Recent Traffic Measurements from an Internet Backbone. 1998.
[16] Chuck Fraleigh et al. Packet-Level Traffic Measurements from the Sprint IP Backbone. IEEE Network, November/December 2003.V.17:P6.
[17] Yeh E, Chao H, Mannem V, Gervais J, Booth B. Introduction to TCP/IP offload engine. 2002. http://www.10gea.org/ SP0502IntroToTOE_F.pdf.
[18] Jon Postel. Transmission Control Protocol – DARPA Internet Program Protocol Specification, RFC 793. September 1981.
[19] John Nagle. Congestion control in IP/TCP internetworks, RFC 896. January 1984
[20] Tanenbaum, A.S. Computer Networks. 3rd ed., Prentice Hall, Inc., 1996.
[21] Peterson, L.L., Davie, B.S. Computer Networks: a System Approach. Morgan Kaufmann Publishers, 2000.
[22] Jacobson, V. Congestion avoidance and control. ACM Computer Communication Review, 1988,18(4):314~329.
[23] Stevens, W. TCP Slow Start, Congestion Avoidance, Fast Retransmit, and Fast Recovery Algorithms. RFC 2001, 1997.
[24] Floyd, S., Henderson, T. The NewReno Modification to TCP’s Fast Recovery Algorithm. RFC 2582, 1999.
[25] L.S. Brakmo, S.W. O'Malley, and L.L. Peterson. TCP Vegas: New techniques for congestion detection and avoidance. 1994 ACM SIGCOMM Conference, pages 24--35, May 1994
[26] Chiu, Dah-Ming, Jain, R. Analysis of the increase and decrease algorithms for congestion avoidance in computer networks. Computer Networks and ISDN Systems, 1989,17(1):1~14.
[27] M. Mathis, J. Mahdavi, S. Floyd, A. Romanow. TCP Selective Acknowledgment Options. RFC 2018, 1996.
[28] S. Floyd, J. Mahdavi, M. Mathis, M. Podolsky. An Extension to the Selective Acknowledgement (SACK) Option for TCP. RFC 2883, 2000
[29] H. Inamura, Ed, G. Montenegro, Ed, R. Ludwig, A. Gurtov, F. Khafizov, TCP over Second (2.5G) and Third (3G) Generation Wireless Networks. RFC 3481,February 2003
[30] Ajay Bakre and B.R. Badrinath. I-TCP: Indirect TCP for Mobile Hosts. In Proceedings of the 15th International Conference on Distributed Computing Systems, 1995.
[31] Microsoft TechNet 中文网站主页. TCP/IP 数据包处理路径. 2005-06
[32] 朱雁辉.Windows 防火墙与数据封包截获技术.北京:电子工业出版社,2002.
[33] Postel, J., Editor.Internet Protocol.STD 5.RFC 791.September 1981.
[34] RFC1072 TCP Extensions for Long-Delay Paths V. Jacobson R. Braden October 1988.
[35] S. Floyd. Limited Slow-Start for TCP with Large Congestion Windows. RFC 3742, 2004
[36] Shacham, A., Monsour, R., Pereira, R., and M. Thomas.IP Payload Compression Protocol (IPComp) .RFC 2393 .December 1998.
[37] Reynolds, J. .J. Postel.Assigned Numbers.STD 2, RFC 1700, October 1994.
[38] P. Deutsch. P.DEFLATE Compressed Data Format Specification version 1.3. RFC 1951, May 1996
[39] RFC 3452 - Forward Error Correction (FEC) Building Block, M. Luby,L. Vicisano,J. Gemmell,L. Rizzo,Univ. Pisa,M. Handley,J. Crowcroft, December 2002.
[40] Dhiman Barman, Ibrahim Matta, Eitan Altman, and Rachid El Azouzi. TCP Optimization through FEC, ARQ and Transmission Power Tradeoffs. Technical report, Boston University, Computer Science Department, Boston, MA 02215, 2003.
[41] Blahut.R.E. .Theory and Practice of Error Control Codes. Addison Wesley.MA.1984.
[42] 北京泰策科技有限公司. 应用层性能测试工具 IxChariot 实验室介绍. 北京,2003.
[43] 北京泰策科技有限公司. 广域网仿真产品 Storm 介绍. 北京,2003
[44] M. Degermark, B. Nordgren, S. Pink. IP Header Compression. RFC2507,1999.
[45] Internet2. NetFlow. Weekly Report. 2007-02-12
[2] InfoWorld.Solving the Branch Office Dilemma:Shifting from Tactical to Strategic Investment,2005-10.
[3] The Strategy Group.Avoiding the Branch Office Quagmire,2005-08.
[4] Cerf,V.,Kahn,R.A protocol for Packet Network Interconnection. IEEE Transactions on Communications. 1974.
[5] I.113. Vocabulary of Terms for Broadband Aspects of ISDN. CCITT. 1988.
[6] I.121. Broadband Aspect of ISDN. CCITT. 1988.
[7] R. Braden, D. Clark, S. Shenker. Integrated Service in the Internet Architecture: An overview. RFC 1633, 1994.
[8] L.Zhang, S. Berson, S. Herzog, S. Jamin. Resource ReSerVation Protocol (RSVP) -Version1 Functional Specification. RFC 2205, 1997.
[9] S. Blake,D. Black, M. Carlson,E. Davies,Z. Wang,W. Weiss. An Architecture for Differentiated Service. RFC 2475, 1998
[10] William Stallings. High-Speed Networks and Internet――Performance and Quanlity of Service, 齐望东,薛卫娟,傅麒麟,胡谷雨等译,谢希仁审校. 北京:电子工业出版社,2003. P239~240
[11] ITU-T. Recommendation Y.1541. Network performance objectives for IP-based services,2002-05:P7
[12] Ceorge Coulourisv,Jean Dollimore,Tim Kinderg. Distributed Systems Concepts and Design Third Edition, 金蓓弘等译. 北京:机械工业出版社,中信出版社,2004. 12~18.
[13] Mark Fabbi,Latency: The Silent Killer of Application Performance. 2004-9-14.
[14] Shunra Software Ltd,The Network Impact: Extending load testing to ensure application performance for remote end users. 2004.
[15] K Claffy,G. Miller,K. Thompson. The Nature of the Beast: Recent Traffic Measurements from an Internet Backbone. 1998.
[16] Chuck Fraleigh et al. Packet-Level Traffic Measurements from the Sprint IP Backbone. IEEE Network, November/December 2003.V.17:P6.
[17] Yeh E, Chao H, Mannem V, Gervais J, Booth B. Introduction to TCP/IP offload engine. 2002. http://www.10gea.org/ SP0502IntroToTOE_F.pdf.
[18] Jon Postel. Transmission Control Protocol – DARPA Internet Program Protocol Specification, RFC 793. September 1981.
[19] John Nagle. Congestion control in IP/TCP internetworks, RFC 896. January 1984
[20] Tanenbaum, A.S. Computer Networks. 3rd ed., Prentice Hall, Inc., 1996.
[21] Peterson, L.L., Davie, B.S. Computer Networks: a System Approach. Morgan Kaufmann Publishers, 2000.
[22] Jacobson, V. Congestion avoidance and control. ACM Computer Communication Review, 1988,18(4):314~329.
[23] Stevens, W. TCP Slow Start, Congestion Avoidance, Fast Retransmit, and Fast Recovery Algorithms. RFC 2001, 1997.
[24] Floyd, S., Henderson, T. The NewReno Modification to TCP’s Fast Recovery Algorithm. RFC 2582, 1999.
[25] L.S. Brakmo, S.W. O'Malley, and L.L. Peterson. TCP Vegas: New techniques for congestion detection and avoidance. 1994 ACM SIGCOMM Conference, pages 24--35, May 1994
[26] Chiu, Dah-Ming, Jain, R. Analysis of the increase and decrease algorithms for congestion avoidance in computer networks. Computer Networks and ISDN Systems, 1989,17(1):1~14.
[27] M. Mathis, J. Mahdavi, S. Floyd, A. Romanow. TCP Selective Acknowledgment Options. RFC 2018, 1996.
[28] S. Floyd, J. Mahdavi, M. Mathis, M. Podolsky. An Extension to the Selective Acknowledgement (SACK) Option for TCP. RFC 2883, 2000
[29] H. Inamura, Ed, G. Montenegro, Ed, R. Ludwig, A. Gurtov, F. Khafizov, TCP over Second (2.5G) and Third (3G) Generation Wireless Networks. RFC 3481,February 2003
[30] Ajay Bakre and B.R. Badrinath. I-TCP: Indirect TCP for Mobile Hosts. In Proceedings of the 15th International Conference on Distributed Computing Systems, 1995.
[31] Microsoft TechNet 中文网站主页. TCP/IP 数据包处理路径. 2005-06
[32] 朱雁辉.Windows 防火墙与数据封包截获技术.北京:电子工业出版社,2002.
[33] Postel, J., Editor.Internet Protocol.STD 5.RFC 791.September 1981.
[34] RFC1072 TCP Extensions for Long-Delay Paths V. Jacobson R. Braden October 1988.
[35] S. Floyd. Limited Slow-Start for TCP with Large Congestion Windows. RFC 3742, 2004
[36] Shacham, A., Monsour, R., Pereira, R., and M. Thomas.IP Payload Compression Protocol (IPComp) .RFC 2393 .December 1998.
[37] Reynolds, J. .J. Postel.Assigned Numbers.STD 2, RFC 1700, October 1994.
[38] P. Deutsch. P.DEFLATE Compressed Data Format Specification version 1.3. RFC 1951, May 1996
[39] RFC 3452 - Forward Error Correction (FEC) Building Block, M. Luby,L. Vicisano,J. Gemmell,L. Rizzo,Univ. Pisa,M. Handley,J. Crowcroft, December 2002.
[40] Dhiman Barman, Ibrahim Matta, Eitan Altman, and Rachid El Azouzi. TCP Optimization through FEC, ARQ and Transmission Power Tradeoffs. Technical report, Boston University, Computer Science Department, Boston, MA 02215, 2003.
[41] Blahut.R.E. .Theory and Practice of Error Control Codes. Addison Wesley.MA.1984.
[42] 北京泰策科技有限公司. 应用层性能测试工具 IxChariot 实验室介绍. 北京,2003.
[43] 北京泰策科技有限公司. 广域网仿真产品 Storm 介绍. 北京,2003
[44] M. Degermark, B. Nordgren, S. Pink. IP Header Compression. RFC2507,1999.
[45] Internet2. NetFlow. Weekly Report. 2007-02-12