Resilient Strategies to SDN: An Approach Focused on Actively Replicated Controllers

Abstract

Software Defined Networking (SDN) are based on the separation of control and data planes. The SDN controller, although logically centralized, should be effectively distributed for high availability. Since the specification of OpenFlow 1.2, there are new features that allow the switches to communicate with multiple controllers that can play different roles -master, slave, and equal. However, these roles alone are not sufficient to guarantee a resilient control plane and the actual implementation remains an open challenge for SDN designers. In this paper, we explore the OpenFlow roles for the design of resilient SDN architectures relying on multi-controllers. As a proof of concept, a strategy of active replication was implemented in the Ryu controller, using the OpenReplica service to ensure consistent state among the distributed controllers. The prototype was tested with commodity RouterBoards/MikroTik switches and evaluated for latency in failure recovery and switch migration for different workloads. We observe a set of trade-offs in real experiments with varyin workloads at both the data and control plane. Resumo. As Redes Definidas por Software (SDN) separam os planos de dados e de controle. Embora o controlador seja logicamente centralizado, ele deve ser efetivamente distribuído para garantir alta disponibilidade. Desde a especificação OpenFlow 1.2, há novas funcionalidades que permitem aos elementos da rede se comunicarem com múltiplos controladores, que podem assumir diferentes papéis -master, slave, e equal. Entretanto, esses papéis não são suficientes para garantir resiliência no plano de controle, pois delega-se aos projetistas de redes SDN a responsabilidade por essa implementação. Neste artigo, exploramos os papéis definidos no protocolo OpenFlow no projeto de arquiteturas resilientes SDN com base em multi-controladores. Como prova de conceito uma estratégia de replicação ativa foi implementada no controlador Ryu usando o serviço OpenReplica para garantir a consistência dos estados. O protótipo foi testado com switches RouterBoards/MikroTik comerciais avaliando-se a latência na recuperação de falha e na migração de switches Agradecimentosà CAPES, CNPq e FAPEMIG pelo suporteà execução deste trabalho.

HawkFlow: Scheme for Scalable Hierarchically Distributed Control in Software Defined Network

“…NIB is designed to adopt mature distributed control system solutions, so it may be faced with problems like poor performance and network state inconsistency. Master/Slaves [14] is proposed to mainly improve the reliability of distributed control plane, working controllers are called Masters, backup controllers are called Slaves, when the tasks of Masters will automatically switched to Slaves when they are detected to be failed. Masters are detected to be failed, but the architecture does not concern the scalability problems of distributed control plane.…”

“…The optimal route of step i can be formulated as, (14) and the distributed optimal route can be formulated as Take the simple topology in Fig. 4 as an example, assume that the routing request is d= (S,T,20), it will take 4 steps of CSR in total for the fully distributed routing, in which step 1 is based on G ag1 with UC1 and r 1 =((b,c,3),(3,4),4), step 2 compute CSR based on G ag2 with UC2 and r 2 =((4,e,6), (6,8),8), step 3 compute CSR based on G ag3 with UC3 and r 3 = ((8,h,k,11), (11,12),12), and step 4 is based on G ag4 with UC4, r 4 =(12,l,m).…”