OpenFlow-Based Dynamic Server Cluster Load Balancing with Measurement Support

—In the current cloud computing environment, the size of the server cluster in the data center is growing in response to the increasing traffic. Due to the use of multiple replicas in the server cluster to provide the same services, effective load balancing as a key technology is very important. In this paper we implement and evaluate an alternative loadbalancing architecture using OpenFlow switches connected to a controller, which gains high flexibility without additional equipment, and has the potential to be more robust than traditional load balancing approach. The system could measure network and server status in real-time and dynamic set weights of server according to the server's processing capability. Our load balancer installs wildcard rules in the switches proactively to direct requests of large groups of clients without involving the controller which effectively saves the flow table space and reduces the delay of the network. Our implementation uses the OpenFlow controller Floodlight and network emulator Mininet to verify the validity of this algorithm. The preliminary evaluation results demonstrate that our dynamic load balancing scheme is superior to not only the random load balancing algorithm but also the round robin load balancing algorithm.