Gnome Meeting

SD-WAN Is Not WAN SDN

With its ability to create self-healing, self-optimizing networks at lower costs, SDN is causing an evolution in the network world. SDN can help service providers and enterprises deliver more flexible and agile services. How can organizations best achieve this? The terms SD-WAN and WAN SDN are thrown around a lot by vendors and the media.

What exactly do these terms mean? What are the differences between them? Here is a primer. (If you are already an SDN master, then you should look to satisfying the new Service Delivery Normal with SDN Management).

Software-Defined Data Center (SDDC)

First, SDN started in the data center. A Software-Defined Data Center (SDDC) is a data center where the core infrastructure – computing, storage, networking, and even security – is virtualized and delivered as a service. The abstraction, management, and orchestration of resources are carried out by hardware-independent, self-optimizing software capable of making intelligent decisions. This differentiates an SDDC from a traditional data center, where each hosted application is allocated a dedicated set of resources at all times, irrespective of whether the application is running or idle. The gain from SDDC is the creation of an agile and energy efficient data center that is flexible to the requirements of any cloud-based or legacy business application.

Software-Defined WAN or SD-WAN

Software-Defined WAN (SD-WAN) takes this concept to the edge of the network, providing an intelligent way to route critical WAN traffic across expensive MPLS circuits and non-critical traffic across the less expensive Internet. The WAN is usually the most expensive and complex part of an enterprise network. Enterprises that need connectivity between their headquarters, branches and data center(s) rely on WAN offerings such as private lines and MPLS circuits from their service providers. While these options provide stable connectivity, they are also expensive. Thus, network engineers are always seeking to deliver the best performance at the least cost. This is where SD-WAN helps.

Say your organization uses a private line or MPLS circuit for inter-branch connectivity, a broadband connection for Internet, and another circuit as backup, which is not only expensive but also sits idle most of the time. You may have considered implementing a hybrid WAN, bundling all your available connections to maximize throughput while reducing costs. But with this, you also have to deal with complex configurations and everyday maintenance to meet the dynamic performance demands of new and existing applications.

SD-WAN is a technology that makes use of the “software defined” concept (de-coupling the control plane from the data plane) and takes it to the edge of the enterprise network. With SD-WAN, the enterprise can automate the management of multiple connections, such as T1, MPLS, broadband, or even LTE into a bundled transport link.

What differentiates this from the hybrid WAN? The control plane or the “software” part of the SD-WAN uses analytics to intelligently automate the configuration, management, and maintenance of the multiple circuits based on the demands of the application, while the data plane works on commodity hardware.

With the SD-WAN approach, different applications traveling to the same destination are set on paths based on their individual requirements. For example, a mission-critical application might use the stable MPLS or T1 circuit, whereas voice and video heading to the same destination may traverse a less expensive path. The result can be significant cost savings on expensive service provider circuits.

How SD-WAN changes the enterprise WAN – Source: Open Networking User Group

WAN SDN

In contrast to SD-WAN, WAN SDN applies the “software defined” concept to the core network. For service providers, this will be the transport or metro area network. For large enterprises that do not outsource their IP/MPLS network, this will be the WAN. SDN can help provide better Quality of Experience, reduce turn-around times for new projects, deploy more services with better ROI, and achieve better operating efficiency in the network.

While the economic advantages of separating the control plane from the data plane extends to WAN SDN, the biggest gain will come from analytics-driven automation to create intelligent networks. Analytics calculate the resource requirements of an application and ensure that requests can be satisfied without adversely impacting current services. Then, they determine the optimal way to provision resources and instruct SDN controllers to program the network accordingly.

Say an application is performing poorly, SDN analytics with automation can determine in real time the cause of poor performance and make intelligent decisions to get around the issue – maybe by redirecting traffic via a different path, or providing better QoS for the application.

Packet Design focuses on WAN SDN, where the management challenges are more complex than in the data center or even SD-WAN. Our analytics-driven SDN Analytics and Automation Platform provides the always-current network routing models, traffic matrices, performance analytics, and user-specific business policies required for intelligent, real-time orchestration by SDN controllers.

You can learn more about SDN analytics and how WAN SDN can help service providers by listening to our recent Webinar or downloading our SDN white paper.

So, where are you with WAN SDN adoption?

Editorial Staff

Add comment

Follow us

Don't be shy, get in touch. We love meeting interesting people and making new friends.

Most popular

Most discussed