微信服务号

微信订阅号

2025年8月13日 12:57 星期三
首页 > 过刊浏览>2023年第34卷第8期 >3924-3937. DOI:10.13328/j.cnki.jos.006541
PDF HTML阅读 XML下载 导出引用 引用提醒
基于流特征的数据中心非对称流负载均衡方法
作者:
作者简介:

陈中卿(1996-),男,硕士,主要研究领域为软件定义网络,下一代互联网.;李丹丹(1987-),女,博士,副教授,CCF专业会员,主要研究领域为网络安全,密码学.;闪德胜(1963-),男,高级工程师,主要研究领域为网络测量,网络安全.;钱叶魁(1980-),男,博士,教授,主要研究领域为网络测量,网络安全.;谢坤(1984-),男,博士,讲师,主要研究领域为下一代网络,网络资源规划与性能优化,智能路由算法.;黄小红(1979-),女,博士,教授,CCF专业会员,主要研究领域为计算机网络应用,下一代互联网,网络安全.;丛群(1980-),男,硕士,主要研究领域为网络管理.

通讯作者:

钱叶魁,E-mail:qyk1129@163.com;黄小红,E-mail:huangxh@bupt.edu.cn

中图分类号:

TP303

基金项目:

国家重点研发计划(2019YFB1802600)


Asymmetric Flow Load Balancing Method Based on Flow Characteristics in Data Center Network
Author:
  • 摘要
    • | |
  • 访问统计
    • |
  • 参考文献 [28]
    • |
  • 相似文献 [20]
    • | | |
  • 文章评论
    摘要:

    数据中心边界广泛部署的地址转换技术产生的非对称流为负载均衡系统的设计带来了挑战. 为了解决软件负载均衡系统不能充分发挥多核处理器和网卡硬件能力的问题, 提出一种基于流特征的非对称流负载均衡方法. 首先, 分析网卡的数据包散列机制, 提出数据包调度算法, 将数据包调度至预期的CPU核; 然后, 基于会话报文序列的时间与空间特征, 构建大象流识别算法; 最后, 基于识别结果, 提出负载均衡方法. 实验结果表明, 非对称流负载均衡方法可以正确处理非对称流的负载均衡, 平均吞吐率提升约14.5%.

    Abstract:

    The asymmetric flow generated by the widely deployed address translation technology brings challenges to the design of load balancing system. To solve the problem of insufficient use of multi-core processors and network card hardware capabilities by software load balancers, an asymmetric flow load balancing method based on flow characteristics is proposed. Firstly, a data packet dispatching algorithm to dispatch packets into expected CPU core via hardware is proposed. Then, an elephant flow detection algorithm is constructed by analyzing of the temporal and spatial characteristics of packet sequences. Finally, based on detected results, a load balance offloading method is proposed. The experimental results show that, asymmetric flow load balancing method can correctly handle the asymmetric flow. Meanwhile, the average throughput rate increases by 14.5%.

    参考文献
    [1] Silva WJA, Dias KL, Sadok DFH. A performance evaluation of software defined networking load balancers implementations. In: Proc. of the 2017 Int’l Conf. on Information Networking (ICOIN). Da Nang: IEEE, 2017. 132–137.
    [2] Jackson EJ, Walls M, Panda A, Pettit J, Pfaff B, Rajahalme J, Koponen T, Shenker S. Softflow: A middlebox architecture for open vswitch. In: Proc. of the 2016 USENIX Annual Technical Conf. Denver: USENIX Association, 2016. 15–28.
    [3] Zhang WS. Linux virtual server for scalable network services. In: Proc. of the 2000 Ottawa Linux Symp. Ottawa, 2000. 1–10.
    [4] Eisenbud DE, Yi C, Contavalli C, Smith C, Kononov R, Man-Hielscher E, Cilingiroglu A, Cheyney B, Shang WT, Hosein J E. Maglev: A fast and reliable software network load balancer. In: Proc. of the 13th Symp. on Networked Systems Design and Implementation. Lombard: USENIX, 2016. 523–535.
    [5] André F, Gouache S, Le Scouarnec N, Monsifrot A. Don’t share, don’t lock: Large-scale software connection tracking with krononat. In: Proc. of the 2018 USENIX Annual Technical Conf. Boston: USENIX Association, 2018. 453–465.
    [6] Jin MX, Wang CS, Li P, Han ZJ. Survey of load balancing method based on DPDK. In: Proc. of the 4th IEEE Int’l Conf. on Big Data Security on Cloud (BigDataSecurity), the IEEE Int’l Conf. on High Performance and Smart Computing (HPSC) and the IEEE Int’l Conf. on Intelligent Data and Security (IDS). Omaha: IEEE, 2018. 222–224.
    [7] 陈福才, 何威振, 程国振, 霍树民, 周大成. 基于DPDK的内网动态网关关键技术设计. 通信学报, 2020, 41(6): 139–151.
    Chen FC, He WZ, Cheng GZ, Huo SM, Zhou DC. Design of key technologies for intranet dynamic gateway based on DPDK. Journal on Communications, 2020, 41(6): 139–151 (in Chinese with English abstract).
    [8] 李凯, 叶麟, 余翔湛, 胡阳. 基于DPDK的流量动态负载均衡方法. 智能计算机与应用, 2017, 7(4): 85–89, 91.
    Li K, Ye, L, Yu XZ, Hu Y. Traffic dynamic load balancing method based on DPDK. Intelligent Computer and Applications, 2017, 7(4): 85–89, 91 (in Chinese with English abstract).
    [9] Gallenmüller S, Emmerich P, Wohlfart F, Raumer D, Carle G. Comparison of frameworks for high-performance packet IO. In: Proc. of the 2015 ACM/IEEE Symp. on Architectures for Networking and Communications Systems. Oakland: IEEE, 2015. 29–38.
    [10] Durner R, Varasteh A, Stephan M, Machuca CM, Kellerer W. HNLB: Utilizing hardware matching capabilities of NICS for offloading stateful load balancers. In: Proc. of the 2019 IEEE Int’l Conf. on Communications (ICC). Shanghai: IEEE, 2019. 1–7.
    [11] Moon YG, Lee SE, Jamshed MA, Park K. AccelTCP: Accelerating network applications with stateful TCP offloading. In: Proc. of the 17th USENIX Symp. on Networked Systems Design and Implementation. Santa Clara: USENIX Association, 2020. 77–92.
    [12] 吴和生, 王崇骏, 谢俊元. 一种多核环境中无锁的多进程负载均衡会话保持方案. 电子与信息学报, 2013, 35(4): 982–987.
    Wu HS, Wang CJ, Xie JY. A lock-free multi-processing session persistence mechanism for load balancing in multi-core environment. Journal of Electronics and Information Technology, 2013, 35(4): 982–987 (in Chinese with English abstract).
    [13] Zheng P, Narayanan A, Zhang ZL. A closer look at NFV execution models. In: Proc. of the 3rd Asia-Pacific Workshop on Networking 2019. New York: ACM, 2019. 85–91.
    [14] 张莹, 吴和生. 面向多进程负载均衡的Hash算法比较与分析. 计算机工程, 2014, 40(9): 71–76.
    Zhang Y, Wu HS. Comparison and analysis of Hash algorithm for multi-process load balancing. Computer Engineering, 2014, 40(9): 71–76 (in Chinese with English abstract).
    [15] He KQ, Hu CC, Jiang JC, Zhou YC, Liu B. A2C: Anti-attack counters for traffic measurement. In: Proc. of the 2010 IEEE Global Telecommunications Conf. Miami: IEEE, 2010. 1–5.
    [16] Woo S, Park KS. Scalable TCP session monitoring with symmetric receive-side scaling. Technical Report, Daejeon: KAIST, 2012. 163–169.
    [17] 刘敬玲, 黄家玮, 蒋万春, 王建新. 数据中心负载均衡方法研究综述. 软件学报, 2021, 32(2): 300–326. http://www.jos.org.cn/1000-9825/6151.htm
    Liu JL, Huang JW, Jiang WC, Wang JX. Survey on load balancing mechanism in data center. Ruan Jian Xue Bao/Journal of Software, 2021, 32(2): 300–326 (in Chinese with English abstract).http://www.jos.org.cn/1000-9825/6151.htm
    [18] Canadian Institute of Cybersecurity. CSE-CIC-IDS2018 on AWS. https://www.unb.ca/cic/datasets/ids-2018.html (2018-10-08)[2020-08-13].
    [19] Scott DW. Box–muller transformation[J]. WIREs Computational Statistics, 2011, 3(2): 177–179. [doi: 10.1002/wics.148]
    [20] Van Tu N, Yoo JH, Hong JWK. Accelerating virtual network functions with fast-slow path architecture using eXpress data Path[J]. IEEE Transactions on Network and Service Management, 2020, 17(3): 1474–1486. [doi: 10.1109/TNSM.2020.3000255]
    [21] Estan C, Varghese G. New directions in traffic measurement and accounting: Focusing on the elephants, ignoring the mice[J]. ACM Transactions on Computer Systems (TOCS), 2003, 21(3): 270–313. [doi: 10.1145/859716.859719]
    [22] 王宏, 龚正虎. Hits和Holds: 识别大象流的两种算法. 软件学报, 2010, 21(6): 1391–1403. http://www.jos.org.cn/1000-9825/3522.htm
    Wang H, Gong ZH. Hits and Holds: Two algorithms for identifying the elephant flows. Ruan Jian Xue Bao/Journal of Software, 2010, 21(6): 1391–1403 (in Chinese with English abstract). http://www.jos.org.cn/1000-9825/3522.htm
    引证文献
    网友评论
    网友评论
    分享到微博
    发 布
引用本文

陈中卿,李丹丹,闪德胜,钱叶魁,谢坤,黄小红,丛群.基于流特征的数据中心非对称流负载均衡方法.软件学报,2023,34(8):3924-3937

复制
分享
文章指标
  • 点击次数:822
  • 下载次数: 2934
  • HTML阅读次数: 1460
  • 引用次数: 0
历史
  • 收稿日期:2021-01-29
  • 最后修改日期:2021-11-06
  • 在线发布日期: 2022-09-30
  • 出版日期: 2023-08-06
文章二维码
友情链接:中国科学院软件研究所中国计算机学会中国科学院国家自然科学基金委员会CCF数字图书馆Int'l J of Softw Info计算机系统应用
您是第20541035位访问者
版权所有：中国科学院软件研究所 京ICP备05046678号-3
地址：北京市海淀区中关村南四街4号,邮政编码：100190
电话：010-62562563 传真：010-62562533 Email：jos@iscas.ac.cn
技术支持：北京勤云科技发展有限公司

京公网安备 11040202500063号