Design And Implementation Of Fault Detection And Recovery For SDN

With the rapid expansion of the scale of the network and the continuous enrichment of business types,the hierarchical structure of the traditional Internet has made the structure and function of the Internet increasingly complicated.For the problems of high difficulty and scalability of traditional network management,SDN is proposed as a new type of network architecture.SDN decouples the control layer from the data layer.Its centralized control and programmable features solve this problem and inject new momentum into the future of network development.However,SDN is in the initial stage of development and lacks a stable and effective fault detection and recovery mechanism.Currently,SDN fault detection and recovery technologies have problems such as large resource expenditure,high operation difficulty,and low efficiency.Therefore,in order to ensure the robustness and reliability of the SDN network and improve the ability of the SDN network to quickly discover and resolve faults,it is necessary to study and implement a fast fault detection and recovery scheme.The specific study of this paper is as follows:(1)For fault detection in SDN networks,the efficiency is low and problems such as node and link failure cannot be solved at the same time.This article presents a hash storage-based SDN link fault detection scheme.This scheme obtains network topology information through the link discovery protocol.In order to improve the speed and efficiency of fault finding,a topology storage method based on hash structure is designed to store all links of the network topology in a key-value pair manner.Finally,a hash-based fault detection algorithm is presented.Experiments show that this scheme can quickly detect the location of faults in locating links and nodes.(2)Aiming at the problems that the backup path in SDN fault recovery is easy to fail and easily lead to network congestion after recovery,this paper proposes a routing algorithm based on multiple weights.This paper introduces the failure rate and bandwidth into the calculation of the backup path to reduce network congestion caused by the backup path and secondary failure of the backup path.For the problem that the backup path is easy to fail,the algorithm calculates multiple backup paths.When the primary backup path fails to recover,the second backup path is used to complete the restoration.In order to meet the 50 ms fault recovery time requirement proposed by the operator,this paper proposes a combined fault recovery scheme.In the early stage of fault recovery,a backup path is used for recovery,and then the controller recalculates the optimal path to complete the fault recovery.Experiments show that this scheme can greatly reduce the failure rate of the backup path and improve the network operating status after the failure recovery.