On Routing Scalability in Flat SDN Architectures

Abstract

The rigidity of traditional network architectures, with tightly coupled control and data planes, impair their ability to adapt to the dynamic requirements of future application domains, such as the Tactile Internet or Holographic-Type Communications. Software-Defined Networking (SDN) architectures, which provide programmability to configure the network, have the potential to provide the required dynamism. However, given its centralized essence, SDN suffers from scalability issues. Therefore, efforts have been made to propose alternative decentralized solutions, such as the flat distributed SDN architecture. Despite its potential, the real applicability and scalability of decentralized SDN solutions are still open research questions. This paper presents a comparative analysis of the effects of different routing approaches on the scalability of flat distributed SDN architectures. Using the Open Network Operating System (ONOS) as our evaluation architecture, we have studied the tradeoff between routing overhead in the control data plane and inter-controller communications for different degrees of decentralization. We have found that routing applications, which only require control-data plane communication for setting the path, benefit more from decentralization than the ones which utilize inter-controller communications and ensure Quality of Service (QoS). Our findings highlight the need for efficient routing mechanisms to deal with inter-controller overhead while lowering the amount of control-data plane communication.