Software defined networking failure recovery with Flow Table aware and flows classification

Abstract

Software Defined Networking (SDN) emerged to simplify network management and speed up innovation through a clear separation of the control plane from the data plane using OpenFlow standard. Link or switch failure occurs more often. The detect switch has to send a port status message to the controller to update the Flow Table with a new path or switch to a preconfigured backup path to recover from the failure. Most of the existing schemes are focusing on distance or delay in backup path selection. However, traffic flows exhibit variabilities and losses caused by flows also differ with respect to Flow Table utilization. Ternary Content Addressable Memory (TCAM) that stores flow entries in the Flow Table is costly with limited space. It is significant to reduce backup path entries resource consumption through different failure recovery strategies while considering the Flow Table utilization ratio at sampling time, flow classification, and shortest path. This paper devised an algorithm for backup path computation based on the shortest path, flow classification, and Flow Table utilization ratio. The simulation result shows higher throughput of (71.2), lower packet losses (3 %), and round-trip time (3.4 s).