Design and implementation of server cluster dynamic load balancing based on OpenFlow

Abstract

Nowadays, the Internet is flooded with huge traffic, many applications have millions users, a single server is difficult to bear a large number of clients' access, so many application providers will put several servers as a computing unit to provide support for a specific application, usually people will use distributed computing, load balancing technology to complete the work. A typical load balancing technique is to use a dedicated load balancer to forward the client requests to different servers, this technique requires dedicated hardware support, the hardware is expensive, lacks of flexibility and is easy to become a single point failure. There will be a new solution for load balancing with OpenFlow proposed., this paper mainly studies dynamic load balancing technology in the OpenFlow environment, the Controller collected the server running status through the SNMP protocol, and calculated the aggregated load of the severs according to dynamic load balancing scheduling algorithm, the OpenFlow switch will forward the client's request to the server whose aggregated load is smallest, thus minimize the response time of the web server. In the OpenFlow network environment, using this method can brings high flexibility without additional equipment.