Impact of Softwarization in Microservices-based SDN Controller
Conference: European Wireless 2022 - 27th European Wireless Conference
09/19/2022 - 09/21/2022 at Dresden, Germany
Proceedings: European Wireless 2022
Pages: 6Language: englishTyp: PDF
Authors:
Scotece, Domenico; Foschini, Luca (Department of Computer Science and Engineering, University of Bologna, Italy)
Arzo, Sisay Tadesse; Devetsikiotis, Michael (Department of Electrical and Computer Engineering, University of New Mexico, USA)
Bassoli, Riccardo (Quantum Communication Networks (QCNets) Research Group, Technische Universität Dresden, Germany & Deutsche Telekom Chair of Communication Networks, Technische Universität Dresden, Germany & Centre for Tactile Internet with Human-in-the-Loop (CeTI), Cluster of Excellence, Dresden, Germany)
Fitzek, Frank H. P. (Deutsche Telekom Chair of Communication Networks, Technische Universität Dresden, Germany & Centre for Tactile Internet with Human-in-the-Loop (CeTI), Cluster of Excellence, Dresden, Germany)
Abstract:
Software-defined networking decouples control and data plane in softwarized networks. This allows for centralized management of the network. However, complete centralization of the controller’s functions raises the issue of the central point of failure, latency, and scalability. Distributed controller deployment is adopted to optimize scalability and latency problems. However, the existing controllers are monolithic, resulting in code inefficiency for distributed deployment. Recently, microservicesbased SDN solutions have been started and deployed as virtual network functions, enabling flexible deployment. Nonetheless, the softwarization of network functionalities introduces network I/O performance degradation both considering deployments based on virtual machines and containers. This paper first introduces a microservices-based SDN solution based on Ryu SDN Framework. Then investigates whether running Ryu’s network functionalities in a softwarized environment (e.g., virtual machine and container) would have a significant network I/O performance degradation. In particular, this work examines more in deep the Docker Container technology and analyzes its network setups. Multiple measurements were performed locally in a single machine. Our results show a comparison between microservicesbased non-virtualized SDN Controller and the virtualized one, and moreover, show a comparison between different Docker Container network setups.