Twisted network cables

How Can You Go From WAN Theory to SD-WAN Reality?

In the last few months a lot of articles have been written about the WAN in general, and about Software-Defined WANs (SD-WANs) in particular. These articles typically point out the well-known limitations of the current approach to wide area networking and discuss how to apply a variety of techniques, both old and new, to the WAN. While articles like this are informative, they often overlook a critical question: How can a network organization move past doing a paper analysis of a new WAN architecture and implement that architecture in a production network?

Do a trial

Only in extremely rare circumstances would it make sense to do a flash cut of a new WAN architecture. One alternative to making a flash cut is to conduct a lab-based trial. In my experience, a lab-based trial provides number of benefits, identifying how well the solution works and highlighting some of its deficiencies.

Unfortunately, only a small percentage of network organizations have the resources and the orientation to conduct lab trials — another alternative is to implement the new architecture in a small number of branch offices. A a branch office approach will allow the organization to experience the benefits of a lab-trial, and even if a network organization has conducted a lab-based trial of any new solution, it typically finds out more about the solution once it is running in the production network.

One of the keys to making the branch-trial approach successful is to choose the branch offices in such a way that they are representative of all of the company’s branch offices. Another key to success is minimizing risk. One way to do this is to implement the trial in an incremental fashion whereby the new WAN architecture is first implemented in just a very few offices. If the solution works well, the trial can be quickly expanded to include additional offices.

Manage existing contracts

One of the potential benefits of an SD-WAN is that it allows a network organization to significantly reduce their use of MPLS. The implementation of that decision might not be possible in the short term, based on the contract that the organization has with their WAN service provider, since most contracts for WAN services include a Minimum Revenue Commitment (MRC) on the part of the company acquiring the services. If the company significantly reduces their use of MPLS, the company’s spend with the service provider could fall below their MRC, which would result in some form of penalty or other action, such as extending the life of the contract.

The fact that a company isn’t able to significantly reduce their use of MPLS in the short term isn’t necessarily a major problem because, as mentioned, few companies would want to do a flash cut of a new WAN architecture. However, in most cases a company spends more with their service provider than just the contractually-committed MRC. As a result, the company can conduct the type of previously-discussed trial, where it implements the new WAN architecture in a small number of branch offices, and it can still meet its contractual obligations.

Reap the benefits

Conducting a trial of a new WAN architecture in a small number of branch offices will identify whether or not it makes sense to implement that architecture more broadly. In addition, having that information puts a network organization in a position of control when the time comes to renegotiate the contract they have with their WAN service provider. The organization could, for example, decide to implement a new WAN architecture and dramatically reduce their MPLS spend. Alternatively, the organization could use the threat of implementing such a WAN architecture as a means of negotiating significantly lower MPLS pricing. Either way, the organization wins.

About the author
Jim Metzler
Jim has a broad background in the IT industry. This includes serving as a software engineer, an engineering manager for high-speed data services for a major network service provider, a product manager for network hardware, a network manager at two Fortune 500 companies, and the principal of a consulting organization. In addition, Jim has created software tools for designing customer networks for a major network service provider and directed and performed market research at a major industry analyst firm. Jim’s current interests include both cloud networking and application and service delivery. Jim has a Ph.D. in Mathematics from Boston University.