Fast Failure Recovery in Software-Defined Networks

In Software defined networks (SDN), the control plane and the data plane are separated. One or more controllers can program a router or switch directly. In an in-band SDN network, the control flow (flow to or out of the controller) and the data stream share a transmission path. Thus, when the fault occurs on the path of the data stream, both the control flow and the data stream are affected. In order to achieve the operator's transmission quality requirements, the network should be able to recover within 50ms. This paper studies the fast fault recovery mechanism in SDN networks. For the control flow and data flow, we have studied the corresponding response and active recovery mechanism, and carried out the simulation experiment. The simulation results show that the reactive recovery mechanism can’t guarantee the failure to recover within 50ms, and the response mechanism of the control plane can cause the delay of data plane fault recovery. The simulation results also show that active recovery of control flow and data flow can meet the requirements of operator network recovery, even in large-scale networks with more traffic.