Loop-Free Rerouting of Multiple Policies With Overlaps

Loop freedom is a fundamental consistency property to ensure that no packets will fall into forwarding loops during network updates. Prior work focuses on tackling the problem of single policy update. However, though some work states that rerouting of policies with overlaps is NP-hard, none of them gave a feasible solution for it. In this paper, we investigate the inherent characteristics of multiple policy cycles (MPCs), and propose general principles to eliminate MPCs. An algorithm named CBreaker is proposed to find the optimal number of rounds to schedule the loop-free update problem. Both theoretical analysis and experimental evaluation demonstrate the feasibility and efficiency of our proposed solution.