Wednesday, August 24, 2011

Big Switch Advances its OpenFlow Platform

Big Switch Networks, a start-up based in Palo Alto, California, reached a significant milestone in its bid to build an open, fully-programmable enterprise switching platform based on Open Flow. The Big Switch Controller has now moved into private beta testing.

OpenFlow, a new network protocol pioneered at Stanford University that enables the path of network packets through a series of switches to be determined by software running on a separate server. This network control plane would managed virtualized switching elements in the network. Virtualized programmable network could open the door to various high-performance applications in campus or enterprise networks. The goal of OpenFlow is to enable networks to evolve more rapidly by giving owners and operators better control over their networks and the ability to optimize network behavior.

Big Switch Networks' Controller is based on OpenFlow and virtualizes the network in the enterprise data center, making it possible for an enterprise to overlay virtual networks on top of their physical devices. The company said its three tier architecture comprises:

1. Big Switch Networks' Controller – a software platform that implements the server side of the OpenFlow protocol, virtualizes the physical resources in the network and provides common components used across OpenFlow Applications

2. OpenFlow Applications – network control applications built on an the Controller's open API that represent a tenant's software-defined devices and features, e.g. virtual switches, virtual firewalls or connectors to server virtualization software

3. Switch Vendors' OpenFlow-Enabled Switches – Ethernet switches, both physical and hypervisor-based, that represent the data plane of an OpenFlow architecture by implementing the client side of the OpenFlow protocol.

  • Big Switch is headed by Guido Appenzeller (Co-Founder and CEO), who previously was a Consulting Professor at Stanford University and Head of the Clean Slate Lab where he managed the OpenFlow standards effort and led the team that developed the OpenFlow reference implementation.