SDN controller distribution method, SDN controller multicast group and network system

Abstract

The present invention discloses an SDN controller distribution method, an SDN controller multicast group and a network system. The method comprises: the SDN control multicast group receives a controller request sent by an OpenFlow switch and sends the controller request to the controller in the SDN controller multicast group, wherein the SDN controller multicast group is formed in the controller cluster mode; the controller in the SDN controller multicast group responds to the controller request according to the load value itself; and the OpenFlow switch selects the controller according to the response condition of the controller in the SDN controller multicast and the load value. The SDN controller distribution method, the SDN controller multicast group and the network system are configured to add controllers into the same multicast group in the cluster mode, and the controller automatic distribution realized based on the multicast mechanism can comprehensively consider the link condition and the load condition when the switch registers the controllers so as to realize the automatic distribution of the cluster controllers which is in zero contact with operation and maintenance staffs.

Description

technical field [0001] The invention relates to the field of SDN networks, in particular to an SDN controller allocation method, an SDN controller multicast group and a network system. Background technique [0002] The controller is the core of the SDN (Software Defined Network, software-defined network) network, responsible for maintaining the status of the global network and calculating routing. The network scale that a single controller can support is limited. The transmission distance between switches is too long, resulting in high network delay and poor performance. Using distributed technology to build a controller cluster, and multiple controllers to manage and control the underlying network collaboratively by region can improve network performance, reliability and scalability. [0003] In the prior art, control optimization of multi-controllers is realized through manual static configuration of switches. In the controller cluster of the master control mode, the mas...

AI Technical Summary

Problems solved by technology

However, static configuration needs to introduce a lot of manual operations, lacks flexibility, and cannot automatically adapt to network changes; the method specified by the master controller will cause a huge load on the master controller, may cause link congestion in the master controller, and does not take into account the control The link between the controller and the switch cannot form the optimal choice; the virtual IP and redirection technology also does not consider the link between the controller and the switch, and cannot form the optimal choice

Images