SDN architectures

Defining Use Cases & Business Cases for SDN

In May 2015, Heavy Reading conducted an operator survey on behalf of Brocade on the topic of network transformation. SDN migration was among the topics covered, and the survey yielded some interesting results. We had 106 qualified responses from the global survey, with 59% coming from the US and 41% from other countries.

Defining use cases has been one of the hottest topics in carrier SDN this year, and we asked survey takers to identify the first use case for their networks, giving them a list of options from which to choose. Network virtualization was the big winner, with nearly one third of the votes. Virtual private cloud and bandwidth on demand formed a second tier, with each receiving 18%. The full results are shown in the chart below.

First Use Cases for SDN
Source: Heavy Reading's May 2015 Network Transformation Survey, sponsored by Brocade; n=105
Source: Heavy Reading's May 2015 Network Transformation Survey, sponsored by Brocade; n=105

Network virtualization creates an abstracted virtual network on top of the existing physical network, and applications can be broad. Typically, general-purpose (or COTS) hardware is used for the physical infrastructure, thus lowering costs. Examples can include creating network functions as a service, creating virtual networks for enterprise branch offices and data center virtual networks.

These results underscore the very tight relationship we are seeing between SDN and NFV. Past Heavy Reading surveys have shown that many network operators view SDN and NFV as interlinked and interdependent, and are building their SDN and NFV products and strategies at the same time.

We also received interesting results when we asked respondents to identify the biggest obstacle to deploying SDN. Lack of a clearly defined business case topped the list (at 28%), followed closely by projected costs of deployment (at 27%).

In the early days of SDN, much of the discussion focused on dramatically lowering capex costs by unseating incumbent suppliers and moving to commodity hardware. That value proposition lost momentum over the past two years, as the other costs of deploying SDN became apparent and as operators realized there are many functions that still require custom hardware. Another issue we have been increasingly hearing about is problematic software licensing models presented by vendors, and we suspect that this issue is reflected in the results.

More surprising to us is the view that SDN business cases are not defined. There has been a tremendous amount of work across the industry in defining and developing use cases for both SDN and NFV, and it continues to be the number one topic of interest in our own discussions with service providers.

However, these survey results show that, despite the progress in focusing on the deployment aspects of the technologies, not enough is being done to prove them in. Internal interest is high among providers, products abound, and well-defined use cases are proliferating, but in many cases the economics of commercialization are yet to be demonstrated. This is an area that will need much greater focus if SDN and NFV are to be successful.

— Sterling Perrin, Senior Analyst, Heavy Reading

This blog is sponsored by Brocade.

Sign In