Research On OpenFlow Flow Rule Processing Performance In Software-Defined Networking

Due to the support of enormous network protocols and a large number of all kinds of network equipment,the network becomes more and more complicated,it lead to the decline of some function.Not only does it limit the IP network technology development,will not be able to meet the current application trend of cloud computing,big data and server virtualization.As a new network paradigm,Software-Defined Networking(SDN)is currently seen as one of the promising approaches in the way towards future Internet,it is proposed to solve the shortcomings of the traditional network.It decoupling control logic from data forwarding,Redefinition and abstraction the control layer,transmission layer and application layer function of network equipment,makes the software programmable of network equipment become possible,and expands the flexibility and programmability of networks,becomes a hot topic in the area of future networks.SDN as a latest network norm making network device programmable with the abstraction redefinition of control plane is expected to solve above issues.This paper first introduced the background,characteristics and development status of SDN;then introduced the status of the dynamic performance of software defined network flue rules processing in detail;and then studied the generating and sending mode of flue rules separately,proposes a solution for the problem of OpenFlow flue rules' processing;finally,summarized this paper and discuss the future research trends.The studies of this paper are as follows:(1)Aiming at the problem of flow rules' generating in software defined networking,we propose a capacity-limited performance evaluation model of SDN controllers M/M/1/m based on queuing theory.Flow rule is occurred in the controller,SDN will come to face the challenge of controller performance bottleneck in its practical deployments.Therefore,it is necessary to understand the performance characteristics of SDN controllers.For this end,we first investigate into the arrival process and the processing time of Packet-in messages in SDN controllers.Then,we propose a capacity-limited performance evaluation model of SDN controllers M/M/1/m based on queuing theory.As a further step,we derive its performance parameters,including the average waiting queue length,the average queue time,the average queue length and the average sojourn time.Finally,our proposed performance model is evaluated by experiments with the controller performance measurement tool Cbench.Experimental results indicate that the estimated delays of our model are closer to the measured delays than those of other existing models.Thus,we can conclude that our proposed model provides a more accurate description of SDN controller performance.(2)Counter the inconsistency in flow table update in software defined networking,we proposes a consistent update algorithm of SDN flow rules based on delay constraints for the in-band control scenario of SDN deployments.A SDN controller introduces delay constraints into the handing out of flow rules to its subordinate switches,and resolves the delay differences between the switches by using shortest path and critical path algorithms,to ensure that network data in a flow can be consistently forwarded by switches in the flow path.Simulation experiment results indicate that the proposed algorithm can guarantee the consistency of flow rule updating,and effectively improve the accuracy of network data forwarding.