IHTA: Dynamic Idle-Hard Timeout Allocation Algorithm based OpenFlow Switch

Abstract

Software Defined Networking (SDN) enables flexible fine-grained networking policies by allowing the SDN controller to install packet handling rules on distributed switches. The behavior of the general network is at the OpenFlow switch which strongly depends on the set of forwarding entries installed at the switch flow table. As the number of traffic flows increase, the number of flow entries augments. Unfortunately, the current switch flow table memory (TCAM) cannot accommodate the number of required entries and eventually switch begins to evict entries from the flow table. Consequently, throttle flow set up request thereby increased the processing load on the controller. Thus, cause the SDN controller scalability problem which decreases the number of switches that could be managed. Since the flow setup requests are generated due to inappropriate timeout setting and premature eviction of flow entries. This paper addresses the SDN scalability problem and proposes a timeout algorithm called idle-hard timeout allocation (IHTA). The algorithm combines both timeouts dynamically according to the traffic pattern, based on Packet inter-arrival time. Therefore, the IHTA timeout dynamically allocates both timeouts with different values to different traffic flows. Flow entry is evicted when they have no packet expected at a sample time. As a result, the experimental result improved the SDN controller scalability problem thereby reducing excessive flow set-up requests. The benchmark results show that the proposed algorithm reduces the number of packet-in message to 35.2% on average.