Preventive Priority Setting Against Multiple Controller Failures in Software Defined Networks

This paper proposes a preventive priority setting model to minimize the worst-case maximum utilization ratio against multiple controller failures in software defined networks. For a set of controllers able to manage a switch, we introduce a priority for each controller to become the main controller that controls the switch. Once the existing main controller fails, the survived controller which has the highest priority works as the main controller. This reassignment is automatically obtained according to the priority setting decided at the network operation start time. In this way, the proposed model provides a prompt recovery with reducing the network instability due to unnecessary controller reassignment. We formulate the proposed model in two different forms, which are an integer linear programming formulation and a min-max formulation. We prove that the considered problem is NP-hard. A basic heuristic is introduced based on the min-max formulation. Its two extensions are further developed considering the accuracy and the computation time in practical computation. Numerical results reveal that, compared to two baselines that sacrifice certain network stability to achieve a more flexible reassignment, the proposed model reduces the network instability by 27% and 52% in average, respectively, with obtaining the comparable maximum utilization ratio.