Evolving from SD-WAN to SD-WAN Edge Solutions

The Next Generation of WAN Solutions

Over the last 18 months there has been a large amount written about Software-Defined WANs (SD-WANs). The vast majority of what has been written has focused on providing more agile, cost-effective transmission services to connect an organization’s branch offices with their internal data centers and with public cloud facilities. I look at the solutions currently in the market, particularly those that focus primarily on providing basic wireline connectivity, as being the first generation of SD-WANs. I believe that the next generation of these solutions will broaden in scope from being an SD-WAN solution to being an SD-WAN edge solution.

The Three Dimensional WAN Challenge

As noted, the vast majority of the first generation of SD-WAN solutions focused primarily on providing connectivity between users in branch offices and the resources that they need to access. While some of those resources reside in internal data centers, a major driver of the adoption of SD-WAN solutions is supporting the large and growing use of public cloud facilities. While the first generation of SD-WAN functionality is very important, it only addresses one dimension of a three-dimensional WAN challenge. The second dimension of the WAN challenge is that IT organizations also need to provide connectivity to other end points at the WAN edge. This includes small/home offices, mobile workers, and the Internet of Things (IoT).

The third dimension of the WAN challenge stems from the fact that providing basic connectivity to all of these end points is insufficient. Network organizations also need to be able to leverage all available bandwidth and utilize as much relatively low-cost consumer broadband as possible. This means that the SD-WAN edge solution must be able to bond all WAN connectivity together in order to satisfy the SLAs and to optimize the user’s experience and productivity. In addition, IT organizations also need to provide a broad set of network functions, and must do that in a way that reduces the complexity found in contemporary networks. A complex network is not an agile network. In addition, operating a complex network consumes resources and doesn’t provide any value that is visible to the company’s business managers.

The network functionality that must be provided includes routing, monitoring and management, WAN optimization, security, load balancing, the ability to set up SSL/IPSec VPNs, and the ability to provide key support services such as DNS and DHCP. The breadth of functionality that is required varies by type of end point. Supporting the IoT may only require security and relatively simple monitoring functionality while supporting branch offices requires a much broader and more sophisticated set of functionality. In order to reduce complexity and cost while increasing agility, as much as possible of the required functionality must be software-based.

The Balance Between Drivers and Enablers

When assessing the emergence of any new class of solutions you have to look for a balance between drivers and enablers. If the drivers are there, but the enablers are not, it could be years before new solutions emerge. Conversely, if the drivers aren’t there but the enablers are, that is technology looking for a solution and in virtually all cases, those solutions die on the vine.

The drivers of the emerging SD-WAN edge solutions are the previously discussed three dimensions of the WAN challenge. While the strength of these drivers varies somewhat by industry and by company, it is difficult to find a company that isn’t impacted by them today and which won’t be impacted by them more heavily in the near term.

One of the key enablers of the emerging SD-WAN edge solutions is the ongoing virtualization of network functionality. This is not a brand new phenomena. We have, for example, had virtualized WAN optimization controllers for several years. What is different now is that because of the interest in Network Functions Virtualization (NFV), a tremendous amount of resource has focused on producing a range of virtualized network functions (VNFs) that perform well without the assistance of dedicated hardware. In addition, significant resources have also focused on related issues, such as how to dynamically service chain virtualized and physical network functions into a complete solution.

Reality Check and Call to Action

SD-WAN solutions are the result of applying an SDN-based approach to connecting an organization’s branch offices with their internal data centers and with public cloud facilities. Expanding on that, the emerging set of SD-WAN edge solutions will feature the integration of a combined SDN- and NFV-based approach in order to fully support all of the end points that connect to the WAN.

It isn’t likely that there will be many fully featured SD-WAN edge solutions available in the market in the next few months. Also, it isn’t reasonable to expect that in the foreseeable future any one vendor on their own will create an SD-WAN edge solution that effectively and efficiently supports all classes of WAN end points and supports all the required network functionality.

In a previous blog I discussed the growing interest in using managed SD-WAN services and suggested that network organizations that are analyzing SD-WANs should consider such solutions. I am also suggesting that network organizations that are analyzing SD-WAN solutions look beyond just providing agile and cost-effective transmission services to branch offices. These organizations should work with vendors to understand the breadth of SD-WAN edge functionality that is currently provided and how that functionality will evolve over time.

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.