As the proliferation of data continues, traditional networking infrastructure is struggling to keep up with the world's influx of information and data. Legacy networking technologies are in need of an overhaul and software-defined networking (SDN) has been touted as the response and the next big thing in virtual networking. Many IT managers already have the flexibility, scalability and automation capabilities of virtualized servers and storage, and want the same attributes in their networks. As businesses enter the second half of 2013, CIOs and CTOs can attest to the serious hype around SDN as the new go-to solution. Many companies - from Arista, VMware (Nicera) to Cisco - have jumped on the bandwagon with their recently developed SDN products and technology roadmaps.

But with every "crash diet" or technology craze, there are still pointed questions that need to be asked: Is this just the latest hype - or are we in a new "networking" reality?

The Most Obvious: Why Now?
SDN began at places like Stanford and Berkeley as a way to build research networks using overlay. In basic terms, SDN is conventionally talked about as an approach to networking where control is decoupled from hardware and given to a software application called a controller. But the main business driver is the need for automation, flexibility and scale as well as the frustration with command-line interfaces and the need to touch hundreds - if not thousands - of boxes and devices to roll out applications and provision services.

Global enterprises are expressing a true and real interest in SDN solutions. The promise is more affordable, simpler, programmable, interoperable automation of service chains and provisioning. Most recently, Google went public with its description of its SDN WAN build out, helping the company to improve networking manageability, lower costs and to share its data among the tens of thousands of data centers the company currently operates. Now that Google has entered the ring, service providers, Web 2.0 companies and select Fortune 500 enterprises are beginning to test the waters.

The Hype: Key Challenges Businesses Face Now
Some may argue that our world has seen quite a few tech trends that were more hype than help. Going from proof of concept (POC) testing to fully operational networks might take customers two to three years. One limitation of SDN solutions today is that they are switch-centric and don't extend to the server endpoints, let alone the host application interface. If the SDN hype continues, enterprises can expect to see interest in "extending technologies" or products that are capable of extending an SDN control plane from the edge switch through the application interface. Open source user level networking solutions that are available today, such as OpenOnload, already provide applications with direct access to the network and an open control plane, enabling some truly unique capabilities in an SDN solution.

The Reality: What Is Being Done Today?
There is more to SDNs than OpenFlow but many startups are working on OpenFlow solutions. OpenFlow is a communications protocol that provides access to the forwarding plane of a network switch or router over the network. Put simply, OpenFlow allows the path of network packets through the network of switches to be determined by software. Architects behind SDN, from Stanford and various Web 2.0 companies, established the Open Networking Foundation to transform networking through the development and standardization of SDN.

If CIOs and CTOs want to see a world where the SDN lives up to the current hype, a focus on open technologies is critical. To mix metaphors, the brave new world of SDN will not be created in seven days and can only come into existence based on robust, open standards. OpenFlow-enabled networks are only a subset of SDNs. Yet arguably the first step toward building an SDN near term is to deploy OpenFlow-enabled devices. For example, Google used switches based on merchant silicon and software based on the OpenFlow specification.

Build for Today to Benefit Tomorrow
There are also several efforts underway to show the beef. Today, technology leaders at customers and research consortia in the U.S., Europe and Asia are beginning to evaluate vendors in order to build OpenFlow networks for test beds and POC testing. A key goal of these POCs is to demonstrate that a multivendor OpenFlow network can perform under typical business loads.

To put it simply, data center managers need to prove the performance of these software-defined networks to start building the transition to this "open" and brave new world.