Adaptive distributed SDN controllers: Application to Content-Centric Delivery Networks

Abstract

This paper aims to propose a deployment solution for Content-Centric Delivery Networks (C-CDN) based on the SDN technology in an IoT-like environment. To meet today’s network requirements in terms of reliability, scalability and performance and to match the specific IoT needs, the SDN architecture needs to be physically-distributed, but logically-centralized. The key idea is to consider the knowledge plane natively provided by the SDN controllers to build Information-Centric Networking (ICN) applications (e.g. Video distribution services). To achieve this, we put forward an adaptive and continuous consistency model for the distributed SDN controllers in large-scale deployments. More specifically, we propose to turn the standard replication technique into a scalable and intelligent replication strategy following Quorum-replicated consistency. That strategy uses the read and write Quorum parameters as adjustable control knobs for a fine-grained consistency level tuning. The main purpose is to find at run-time appropriate partial Quorum configurations that achieve, under changing network and workload conditions, balanced trade-offs between the SDN application’s continuous performance and consistency requirements. When compared to ONOS’s static consistency model, our approach which was implemented for a Content-Centric Delivery Network (C-CDN) application that we designed on top of ONOS, proved efficient in minimizing the application’s inter-controller overhead, while satisfying the SLA-style application requirements.