A High-Performance and Accurate FPGA-Based Flow Monitor for 100 Gbps Networks

Author:

Sha MengORCID,Guo Zhichuan,Wang KeORCID,Zeng Xuewen

Abstract

Network monitoring is important to improve network performance and security. The separation of the data plane and the control plane of SDN gives the network flexibility. This flexibility facilitates network management and gives rise to the need for accurate and fine-grained management of networks. However, the traditional software-based flow monitoring cannot easily keep up with today’s high-speed networks such as 100 GbE, even with the help of high-performance frameworks such as DPDK. To meet the need for fine-grained management of high-speed networks, an accurate hardware-based flow monitor design is presented in this paper. In our design, the FPGA-based pipelined cuckoo hashing is used to achieve efficient storage of flow entries. The flow information is accurately recorded without any sampling. The proposed design can achieve a higher performance than the non-sketch CPU-based methods and higher accuracy than sketch methods. Compared with other state-of-the-art flow monitors, the proposed design can achieve a performance of 100 Gbps while maintaining high accuracy.

Funder

Strategic Leadership Project of Chinese Academy of Sciences: SEANET Technology Standardization Research System Development

Publisher

MDPI AG

Subject

Electrical and Electronic Engineering,Computer Networks and Communications,Hardware and Architecture,Signal Processing,Control and Systems Engineering

Reference23 articles.

1. Scalable Rule Management for Data Centers;Moshref;Proceedings of the 10th USENIX Symposium on Networked Systems Design and Implementation (NSDI 13),2013

2. Hedera: Dynamic Flow Scheduling for Data Center Networks;Al-Fares;Proceedings of the 7th USENIX Symposium on Networked Systems Design and Implementation (NSDI 10),2010

3. Flow Monitoring Explained: From Packet Capture to Data Analysis With NetFlow and IPFIX

4. Software-defined networking (SDN): a survey

5. Devoflow: Cost-Effective Flow Management for High Performance Enterprise Networks;Mogul;Proceedings of the 9th ACM SIGCOMM Workshop on Hot Topics in Networks,2010

Cited by 3 articles. 订阅此论文施引文献 订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3