Elements of SD-WAN platforms: Differentiators, characteristics and the hidden secret

Published on June 2nd, 2020

One word is beginning to dominate the SD-WAN industry vernacular: platform. SD-WAN solutions have become mashups that squeeze more appliances, features, services, cloud applications and AI tools into one offering, making all of them now platforms. While these SD-WAN medleys might be as refreshing as a bouquet of assorted spring flowers to some, the term platform has become so over-used that it’s losing definition. In other words, the bouquet is wilting.

With platforms everywhere, it’s more important than ever to avoid the ones that leave IT systems fragmented and to invest time finding the natively integrated springboards to success. But to get there, we first need to define what an SD-WAN platform actually is, the characteristics that differentiate platforms from mere solution sets, and the hidden secret that enables users to easily provision, configure, orchestrate, and manage their network in this digital age of constant change.

Defining an SD-WAN platform: Start with the core

Quite literally, a platform is a standard foundation or lowest support substructure from which the larger SD-WAN solution is built. The degree to which the underlying SD-WAN architecture is standardized distinguishes platforms from solution sets. The key is one transparent ecosystem. When all of an SD-WAN’s ancillary capabilities share one common infrastructure from edge to edge, benefits include:

The core is what excludes many of today’s market solutions from being SD-WAN platforms. What lies beneath will largely dictate the customer experience, and, ultimately, success. Managed SD-WAN providers typically will use their own private network to serve as the SD-WAN platform. In these cases, users need to question whether the provider’s network is entirely software defined, and especially inquire whether it was built using a ubiquitous SD-architecture. If the entire foundation is software defined, virtually every network function will be easily programmable and modifiable via software.

A ubiquitous SD-core is the hidden secret. When the provider expands its SD-WAN solution with more appliances, features, services, applications and Artificial Intelligence (AI), it embeds all of those tools into the same SD-network fabric. This is how a platform extends its value – it gives its medley of ancillary capabilities a common ground and governed by software. Unlike solution sets, platforms deploy a vast number of tools into many different client environments – multiple clouds, branch offices, edge networks, mobile devices and data centers – and unite them all under a holistic service and management framework that provides deep visibility and end-to-end, uniform control.

Critical characteristics: Global, cloud-based, AI-powered
Features and applications embedded into the platform should interoperate as equal partners on the same operating system. Furthermore, the platform should also be global and cloud-based. With a platform that has global reach, the SD-WAN solution can make more promises of service consistency into every corner of your enterprise footprint. And with cloud and virtual delivery models, it provides not only the agility and responsiveness you require, but ultimately paves the way for automation.

Much the same way that software-defined network fabrics are the platforms of SD-WAN, AI-based analytics and automation tools are the cornerstones needed to build fully automated networks or the autonomous networks of the future. Which means you want your SD-WAN platform to be leveraging AI-based analytics now, so that one day, it just might be smart enough to be self-optimizing, self-healing and self-driving.

Let’s look at the way advanced SD-WAN platforms are using AI to automate network optimization and lean into the future. First, behavior analytics make sense of the seemingly endless streams of data describing the activities and state of a network. Machine learning turns that mountain of data into comprehensible intelligence, providing a human-understandable picture of what is going on, and makes that data actionable. This new level of understanding can provide:

These capabilities are available today. Network virtual advisors are made possible through AIOps tools (which stands for AI for IT Operations). Essentially AIOps can act as a virtual network assistant working 24/7. But this is just the beginning of autonomous networking.

Platforms: Breeding grounds for intent-based and autonomous networks
When platforms also have automation tools embedded in the SD-network fabric, IT teams have the ability to program them using workflows and playbooks based on their standard network operations. This allows for fully automated network optimization and automated security improvements that will become the intent-based networks and autonomous networking of tomorrow.

Reaching full autonomy is virtually impossible without the purity of the SD-WAN platform’s underlying architecture. The complexity of the management environment affects the use of AIOps. SD-WAN platforms will be unable to provide AIOps tools the information they need to be truly effective when there are more generations of technology, more diverse infrastructures from more vendors and more protocols for integration. The value of AIOps can be made or broken based on the quality of the SD-WAN platform.

The 5 elements of an SD-WAN platform

To recap, here are the five essential elements that make up an SD-WAN platform.

Terry Traina

Terry Traina is the chief technology officer and has been with Masergy since 2004. Terry leads various teams including research and development, platform architecture, software engineering, and product development. Terry has over 20 years of experience in the telecommunications industry specializing in building and managing everything from enterprise software-defined networks to routing and switching platforms. Prior to joining Masergy, Terry has held various technical positions at Chiaro Networks, Allied Riser Communications, and EDS. Terry holds a bachelor’s of science in electrical engineering from University of Southwestern Louisiana.

Related Content

Why are there so many different interpretations of SASE? The many facets of SASE.

Ask people what SASE is and you’ll likely get 10 answers. While the newness of SASE is a factor in the confusion, here’s what SASE is and what it’s not.

Read more

Poised for continued innovation, Masergy welcomes Ajay Pandya as Director of Product Management – Network Solutions

Masergy’s new Director of Networking explains the questions driving ingenuity and the company’s next generation of solutions.

Read more

Enabling 5,500 remote employees in one weekend: How one CIO was already prepared for COVID-19

What if you had started building a work-from-anywhere IT platform before COVID-19? Joe Gracz shares his personal story with the aim of helping others.

Read more