When software defined networks meet fault tolerance: A survey

Abstract

Software Defined Network (SDN) is emerging as a novel network architecture which decouples the control plane from the data plane. However, SDN is unable to survive when facing failure, in particular in large scale data-center networks. Due to the programmability of SDN, mechanism could be designed to achieve fault tolerance. In this survey, we broadly discuss the fault tolerance issue and systematically review the existing methods proposed so far for SDN. Our representation starts from the significant components that OpenFlow and SDN brings - which are useful for the purpose of failure recovery, and is then further expanded to the discussion of fault tolerance in data plane and control plane, in which two phases - detection and recovery - are both needed. In particular, as the important part of this paper, we have highlighted the comparison between two main methods - restoration and protection - for failure recovery. Moreover, future research issues are discussed as well.