SDN Solution for Network Load Balancing

Project Overview

PLVision’s team developed an infrastructure SDN solution for network load balancing in the Customer’s data center. The solution included Data Plane (a scalable cluster of virtual switches), Control Plane (OpenDaylight deployed and customized) and Application Plane (load measurement and balancing algorithms).

Customer Profile

Information and communications technology service provider.

Business Needs Business Needs

  • The Customer had a need to improve server infrastructure utilization, as well as to support virtual network auto scaling capability depending on a number of servers and requests.
  • Changes were required to OpenDaylight UI and API to collect the load of the system, and provide comprehensive interface for management and monitoring.

Technical Highlights Technical Highlights

  • Key technologies: OpenFlow, OpenDaylight, OVS, Python, Java, REST.
  • Existing OpenDaylight UI was extended to support flexible adjustment of all the required parameters (load balancing algorithms, server pools, flow tables etc.).
  • A cluster of SDN switches was deployed to enable network forwarding decisions in order to send service requests from clients of a service to multiple servers in a server pool, which resulted in roughly equal servers’ load balancing.
  • The support of the network auto scaling was implemented, which ensured stable performance during high-traffic events and load changes.

Business Benefits Business Benefits

  • The use of open-source components allowed the Customer to avoid huge investments and decrease the product’s time to market.
  • Development of an MVP with core functionality for demo and presale purposes enabled the Customer to eliminate risks associated with the new product launch, specifically allocation of new investments and impact on the existing business processes.
  • Implementation of OpenDaylight Load Balancer UI provided more options for network configuration and monitoring as well as enabled comprehensive management of the entire network.
  • The solution resulted in better infrastructure utilization and performance as well as enhanced security (through complex filtering rules) and user experience (through better quality of service).