Dynamic monitoring model based on DPDK parallel communication

doi:10.11772/j.issn.1001-9081.2019081405

Journal of Computer Applications ›› 2020, Vol. 40 ›› Issue (2): 335-341.DOI: 10.11772/j.issn.1001-9081.2019081405

• DPCS 2019 • Previous Articles Next Articles

Dynamic monitoring model based on DPDK parallel communication

Cui LI, Qingkui CHEN()

School of Optical-Electrical and Computer Engineering，University of Shanghai for Science and Technology，Shanghai 200093，China

Received:2019-07-31 Revised:2019-09-03 Accepted:2019-09-19 Online:2020-02-26 Published:2020-02-10
Contact: Qingkui CHEN
About author:LI Cui， born in 1994， M. S. candidate. Her research interests include network communication， parallel computing.
Supported by:
the National Natural Science Foundation of China(61572325);the Research Fund for the Doctoral Program of Higher Education of China(20113120110008);the Key Science and Technology Research Project of Shanghai(14511107902);the Construction Project of Shanghai Engineering Center(GCZX14014);the Shanghai Top-class Discipline Construction Project(XTKX2012);the Hujiang Foundation of China(C14001┫．)

基于DPDK并行通信的动态监控模型

李翠, 陈庆奎()

上海理工大学光电信息与计算机工程学院，上海 200093

通讯作者: 陈庆奎
作者简介:李翠（1994—），女，甘肃武威人，硕士研究生，主要研究方向：网络通信、并行计算
基金资助:
国家自然科学基金资助项目(61572325);高等学校博士学科点专项科研博导基金资助项目(20113120110008);上海重点科技攻关项目(14511107902);上海市工程中心建设项目(GCZX14014);上海市一流学科建设项目(XTKX2012);沪江基金研究基地专项(C14001)

Abstract

Abstract:

In order to make better use of the performance of communication system， make full use of the resources of system nodes， and improve the reliability and stability of the system， a dynamic monitoring model based on DPDK （Data Plane Development Kit） parallel communication was designed. Combined with the high speed， large traffic， strong real-time characteristics of DPDK and communication system， the model was designed for multi-node backup， data packet and control packet separation， transmitting and receiving data packets in parallel by multiple network ports， and processing data packets in parallel by multiple cores. The monitoring objects were analyzed， the data acquisition methods were researched， and a 2-layer communication protocol named DMPD was designed. The fine-grained monitoring was performed on the network ports， and the network port load information model was given. Besides， a more efficient and fair dynamic load balancing algorithm based on multi-network port was designed by combining hash function， adjustment function and dynamic load information. Experimental results show that the dynamic monitoring model can accurately detect and timely deal with the abnormality appeared in the system， and achieve the load balancing of multiple network ports.

Key words: DPDK （Data Plane Development Kit）, parallel communication, dynamic monitoring, multi-network port, load balancing

摘要：

为了更好地发挥通信系统的性能，充分利用系统节点的资源，提高系统的可靠性与稳定性，设计了一种基于DPDK并行通信的动态监控模型。该模型结合DPDK和通信系统的高速率、大流量、强实时性等特点，面向多节点备份、数据包与控制包分离、多网口并行收发数据包、多核并行处理数据包进行设计，分析了监控对象，研究了数据采集方法，设计了二层通信协议DMPD，并对网口进行了细粒度监控，给出了网口负载信息模型。另外，将散列函数、调整函数与动态负载信息结合起来设计了更有效、更公平的基于多网口的动态负载均衡算法。实验结果表明，该监控模型能够准确检测和及时处理系统出现的异常，并且实现了多网口的动态负载均衡。

关键词: DPDK, 并行通信, 动态监控, 多网口, 负载均衡

CLC Number:

TP393.1

Cui LI, Qingkui CHEN. Dynamic monitoring model based on DPDK parallel communication[J]. Journal of Computer Applications, 2020, 40(2): 335-341.

李翠, 陈庆奎. 基于DPDK并行通信的动态监控模型[J]. 《计算机应用》唯一官方网站, 2020, 40(2): 335-341.

Figures/Tables 14

References 20

1	BONOMI F， MILITO R， ZHU J， et al. Fog computing and its role in the Internet of things［C］// Proceedings of the 1st Edition of the MCC Workshop on Mobile Cloud Computing. New York： ACM， 2012： 13-16. 10.1145/2342509.2342513
2	NINAGAWA C， IWAHARA T， SUZUKI K. Enhancement of OpenADR communication for flexible fast ADR aggregation using TRAP mechanism of IEEE1888 protocol［C］// Proceedings of the 2015 IEEE International Conference on Industrial Technology. Piscataway： IEEE， 2015： 2450-2454. 10.1109/ICIT.2015.7125458
3	BELOGLAZOV A， ABAWAJY J， BUYYA R. Energy-aware resource allocation heuristics for efficient management of data centers for cloud computing［J］. Future Generation Computer Systems， 2012， 28（5）：755-768. 10.1016/j.future.2011.04.017
4	SHI W， CAO J， ZHANG Q， et al. Edge computing： vision and challenges［J］. IEEE Internet of Things Journal， 2016， 3（5）： 637-646. 10.1109/jiot.2016.2579198
5	HU Y C， PATEL M， SABELLA D， et al. Mobile edge computing — a key technology towards 5G［J］. ETSI White Paper， 2015， 11（11）： 1-16.
6	FD.io. How does fd.io relate to DPDK？［EB/OL］. ［2019-05-16］. . 10.1038/sj.cdd.4400341
7	Intel. Data plane development kit［EB/OL］. ［2019-05-16］. . 10.1201/9780429353512-17
8	DPDK Intel. Programmer’s guide ［EB/OL］. ［2019-05-12］. .
9	LUO T， WANG X， HU J， et al. Improving TLB performance by increasing hugepage ratio［C］// Proceedings of the 15th IEEE/ACM International Symposium on Cluster， Cloud and Grid Computing. Piscataway： IEEE， 2015： 1139-1142. 10.1109/ccgrid.2015.36
10	DPDK Intel. Project charter［EB/OL］. ［2019-05-12］. .
11	CIUFFOLETTI A. Beyond Nagios-design of a cloud monitoring system［C］// Proceedings of the 6th International Conference on Cloud Computing and Services Science. Setúbal： SciTePress， 2016： 363-370. 10.5220/0005778303630370
12	ANUSAS-AMORNKUL T， SANGRAT S. Linux server monitoring and self-healing system using Nagios［C］// Proceedings of the 14th International Conference on Mobile Web and Intelligent Information Systems， LNCS10486. Cham： Springer， 2017： 290-302.
13	KATSAROS G， KÜBERT R， GALLIZO G. Building a service-oriented monitoring framework with REST and Nagios［C］// Proceedings of the 2011 IEEE International Conference on Services Computing. Piscataway： IEEE， 2011： 426-431. 10.1109/scc.2011.53
14	郭晓慧，李润知，张茜，等.基于Zabbix的分布式服务器监控应用研究［J］.通信学报，2013，34（Z2）：94-98.
	GUO X H， LI R Z， ZHANG Q， et al. Application research on distributed Zabbix network monitoring system［J］. Journal on Communications， 2013， 34 （Z2）： 94-98
15	赵哲，谭海波，赵赫，等.基于Zabbix的网络监控系统［J］. 计算机技术与发展，2018，28（1）：144-149. 10.3969/j.issn.1673-629X.2018.01.031
	ZHAO Z， TAN H B， ZHAO H， et al. Network monitoring system based on Zabbix［J］. Computer Technology and Development， 2018， 28 （1）： 144-149. 10.3969/j.issn.1673-629X.2018.01.031
16	CERRATO I， ANNARUMMA M， RISSO F. Supporting fine-grained network functions through Intel DPDK［C］// Proceedings of the 3rd European Workshop on Software Defined Networks. Piscataway： IEEE， 2014： 1-6. 10.1109/ewsdn.2014.33
17	令瑞林，李峻峰，李丹.基于多核平台的高速网络流量实时捕获方法［J］.计算机研究与发展，2017， 54（6）：1300-1313. 10.7544/issn1000-1239.2017.20160823
	LING R L， LI J F， LI D. Realtime capture of high-speed traffic on multi-core platform［J］. Journal of Computer Research and Development， 2017， 54 （6）： 1300-1313. 10.7544/issn1000-1239.2017.20160823
18	DURNER R， VARASTEH A， STEPHAN M， et al. HNLB： utilizing hardware matching capabilities of NICs for offloading stateful load balancers［C］// Proceedings of the 2019 IEEE International Conference on Communications. Piscataway： IEEE， 2019： 1-7. 10.1109/icc.2019.8761434
19	Mircosoft. Receive Side Scaling （RSS）［EB/OL］. ［2019-05-24］. . 10.1145/3359989.3365412
20	WILES K. Pktgen-DPDK［EB/OL］. ［2019-05-15］. .

组件	监控的内容
CPU	CPU的温度、核数目、核利用率、利用率、负载、进程数和线程数
内存	内存大小、内存利用率
网口	网口的数量、流量、丢包率、状态、负载和利用率
电源	工作电压、工作电流
I/O	I/O利用率
网络	网络流量、网络负载
管道	管道的数目、大小、状态和优先级

组件	监控的内容
CPU	CPU的温度、核数目、核利用率、利用率、负载、进程数和线程数
内存	内存大小、内存利用率
网口	网口的数量、流量、丢包率、状态、负载和利用率
电源	工作电压、工作电流
I/O	I/O利用率
网络	网络流量、网络负载
管道	管道的数目、大小、状态和优先级

配置项	配置信息
CPU	Intel Core i7-4790 CPU @3.60 GHz/ Intel Xeon CPU E3-1230 V2@3.30 GHz 8核
操作系统	Centos7.0
内核版本	Linux version 3.10.0-862.14.4.el7.x86_64
网卡	4端口，i350
内存	16 GB/24 GB/32 GB，DDR3
大页	4 GB/8 GB

配置项	配置信息
CPU	Intel Core i7-4790 CPU @3.60 GHz/ Intel Xeon CPU E3-1230 V2@3.30 GHz 8核
操作系统	Centos7.0
内核版本	Linux version 3.10.0-862.14.4.el7.x86_64
网卡	4端口，i350
内存	16 GB/24 GB/32 GB，DDR3
大页	4 GB/8 GB

监控对象	数据包大小/B	核数目	CPU利用率/%	CPU平均负载值			内存/GB	内存占用率/%
监控对象	数据包大小/B	核数目	CPU利用率/%	1 min	5 min	15 min	内存/GB	内存占用率/%
master	>1 024	8	100	8.51	8.22	6.78	16	59
node1	64	8	100	8.28	8.30	6.01	16	57
node2	64	8	100	8.92	8.24	6.85	24	42
node3	128	8	100	7.82	7.50	5.38	24	38
node4	128	8	100	7.87	7.57	5.66	16	56
node9	256	8	100	7.85	7.56	5.42	16	57
node5	512	8	100	8.21	7.92	6.78	16	56
node6	512	8	100	8.16	7.89	6.66	16	58
node7	1 024	8	100	7.98	7.78	5.86	16	57
node8	1 024	8	100	8.07	8.01	6.23	24	37

Dynamic monitoring model based on DPDK parallel communication

基于DPDK并行通信的动态监控模型

RichHTML

PDF

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 14

References 20

Related Articles 15

Recommended Articles

Metrics

监控对象	网口数量	网口状态	网口数据流速率/（MB·s^-1）	网口利用率/%
master	8	1，1，1，1， 1，1，1，1	114，123，124，108， 117，124，103，120	91，98， 99， 86， 93， 99， 82， 96
node1	8	1，0，1，1， 1，1，1，1	90， 0， 77， 90 ， 84，91，71，85	72 ，0， 62， 72， 67，72， 56， 68
node2	4	1，1，1，0	91，103，83，0	72， 82， 66， 0
node3	8	1，1，1，1， 1，1，1，1	108， 100， 107，108， 118， 103， 100， 96	86， 80， 86， 86， 94， 82， 80， 77
node4	4	1，1，1，1	106， 113 ，99， 107	85， 90， 79， 86
node9	4	1，1，1，1	120， 101， 113 ，116	96， 81， 90， 93
node5	4	1，1，1，1	101， 116， 108，121	81， 93， 86， 97
node6	8	1，1，0，1， 1，1，1，1	110， 121， 0，102， 118， 106 ，120，101	88， 97， 0， 82， 94， 85， 96， 81
node7	8	1，1，1，1， 1，1，1，1	105，120， 113 ，118， 124，114 ，117 ，122	84， 96， 90， 94， 99， 91， 94， 97
node8	4	1，1，1，1	111，119， 104 ，123	89， 95， 83， 98

[1]	ZHAO Quan, TANG Xiaochun, ZHU Ziyu, MAO Anqi, LI Zhanhuai. Low-latency cluster scheduling framework for large-scale short-time tasks [J]. Journal of Computer Applications, 2021, 41(8): 2396-2405.
[2]	XU Hongliang, YANG Guiqin, JIANG Zhanjun. Data center adaptive multi-path load balancing algorithm based on software defined network [J]. Journal of Computer Applications, 2021, 41(4): 1160-1164.
[3]	YANG Ling, JIANG Chunmao. Strategy of energy-aware virtual machine migration based on three-way decision [J]. Journal of Computer Applications, 2021, 41(4): 990-998.
[4]	LI Zhuhong, ZHAO Canming, YAN Long, ZHANG Xinming. Load balancing opportunistic routing protocol for power line communication network in smart grids [J]. Journal of Computer Applications, 2019, 39(3): 812-816.
[5]	WANG Zewu, SUN Lei, GUO Songhui, SUN Ruichen. Scheduling method of virtual cipher machine based on entropy weight evaluation in cryptography cloud [J]. Journal of Computer Applications, 2018, 38(5): 1353-1359.
[6]	FU Chaojiang, WANG Tianqi, LIN Yuerong. Parallel algorithm for explicit finite element analysis based on efficient parallel computational strategy [J]. Journal of Computer Applications, 2018, 38(4): 1072-1077.
[7]	LU Liang, YU Jiong, BIAN Chen, YING Changtian, SHI Kangli, PU Yonglin. Task scheduling algorithm based on weight in Storm [J]. Journal of Computer Applications, 2018, 38(3): 699-706.
[8]	GENG Haijun, LIU Jieqi. Distributed load balancing algorithm based on hop-by-hop computing [J]. Journal of Computer Applications, 2018, 38(12): 3524-3528.
[9]	TAI Yingying, PANG Ying, DUAN Keke, FU Yunpeng. Dynamic algorithm of load balancing based on D-S evidence theory with improved weight [J]. Journal of Computer Applications, 2018, 38(10): 2976-2981.
[10]	ZHOU Yue, CHEN Qingkui. Load balancing mechanism for large-scale data access system [J]. Journal of Computer Applications, 2018, 38(1): 50-55.
[11]	ZHOU Weibo, ZHONG Yong, LI Zhendong. Storage load balancing algorithm based on storage entropy [J]. Journal of Computer Applications, 2017, 37(8): 2209-2213.
[12]	LI Qiang, LIU Xiaofeng. Load balancing strategy of cloud storage based on Hopfield neural network [J]. Journal of Computer Applications, 2017, 37(8): 2214-2217.
[13]	HAN Hu, WANG Peng, CHENG Kun, LI Bo. Task scheduling algorithm for cloud computing based on multi-scale quantum harmonic oscillator algorithm [J]. Journal of Computer Applications, 2017, 37(7): 1888-1892.
[14]	LIU Biguo, SHU Yong'an, FU Yinghui. Load balancing scheme based on multi-objective optimization for software defined network [J]. Journal of Computer Applications, 2017, 37(6): 1555-1559.
[15]	YANG Zhongming, LIANG Benlai, QIN Yong, CAI Zhaoquan. Dynamic load balancing algorithm based on monitoring and adjusting of multiple detection engines [J]. Journal of Computer Applications, 2017, 37(3): 717-721.