Impact of Southbound Expansion on Clustered OpenFlow Software-Defined Network Controller Synchronisation Using ODL and ONOS

Author:

Hettiarachchi Egodahettiarachchige Don Sarada Indumini1ORCID,Sarkar Nurul I.1ORCID,Gutierrez Jairo1ORCID

Affiliation:

1. Computer Science and Software Engineering, Auckland University of Technology, Auckland 1010, New Zealand

Abstract

The clustering methods of Software-Defined Networking (SDN) have gained popularity due to their ability to offer improved scalability, consistency, dependability, and load balancing within overlay networks and SDN partitions. This paper delved into the effects of increasing the number of OpenFlow-enabled southbound devices on the establishment and coordination of SDN-controller clusters. Specifically, we examined the volume of east–west cluster packet communications concerning the number of southbound devices within the topology. Many research studies have focused on bandwidth and the number of bytes in east–west communication. While bandwidth refers to the maximum rate at which data can be transferred, and the number of bytes reflects the volume of data being transmitted, the number of packet communications directly influences the efficiency and responsiveness of network operations. Our investigation encompassed the impact of SDN controller-to-controller communication within the cluster concerning the rising number of OpenFlow switches connected to various topologies, including tree (star-bus network), linear, and torus configurations. This study provided data on communication patterns within Open Network Operating Systems (ONOS) and OpenDaylight (ODL) clusters, revealing differing levels of controller communication with southbound network expansions. We evaluated the scalability of ODL and ONOS controllers by scrutinising the effect of increasing the number of southbound devices on the control communication volume. Our analysis revealed varied communication patterns within ONOS and ODL clusters, resulting in different volumes of control communication with southbound expansions. The findings indicated that in small-to-medium-sized SDNs, ODL outperformed ONOS, notably with faster cluster discovery. Conversely, ONOS demonstrated greater efficiency in larger networks owing to its centralised communication architecture. Finally, we provide recommendations for selecting the most suitable controllers based on the size of southbound networks, aiming to provide practical guidelines for optimal network performance.

Publisher

MDPI AG

Reference32 articles.

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