Remember the early years of child rearing when you told your son or daughter that it was time to clean their room, but you had to sit down on the floor and help them pick up all the toys or the work would never get done? As children mature, they grow out of that phase, and they start doing the work without your help and detailed instruction. They start to understand the intent behind your command and the tasks need to achieve the goal.
Well, networks are making similar advances. It’s called intent-based networking.
Masergy ran a quick social media poll to get a pulse on the number of people who have heard of intent-based networking. Only 30% of our respondents are familiar with the term. Clearly, it needs an introduction. Let’s explore the evolution of networking technologies and how to build a self-modifying network.
In 2015, I wrote an article labeling software defined networking (SDN) as a technology on the rise. The article talked about how SDN promised to transform enterprise networks into modular assets that can be assembled and rearranged as business needs require, ultimately making networks just as programmable as the components connected to them.
With all of that in the rearview mirror now, we see that SDN has indeed delivered, as evidenced by the adoption rates in both the data center and wide area networking including SD-WAN. By 2020, Gartner predicts the programmatic capabilities of a software defined data center will be considered a requirement for 75 percent of Global 2000 enterprises that seek to implement a DevOps approach and a hybrid cloud model. SD-WAN is following suit. Gartner blogger Andrew Lerner recently reported, “While many networking technologies are over-hyped as the next big thing, SD-WAN is delivering on the promise. In just three short years, adoption has taken off. . .” Additionally, Gartner’s 2018 Magic Quadrant report for Global Network Services states nearly 40% of its survey respondents are using or evaluating SD-WAN specifically, while another 26% are planning for SD-WAN in the next 18 months.
But beyond just a technology “on the rise,” SDN has become a foundational element for the future of enterprise networking. SDN is paving the way for intent-based, self-sufficient networks.
Software-defined networking infrastructure is a significant advancement because it decouples two primary components of networking: the control plane and the data plane. Previously these two operations were intrinsically linked, but software-defined networks separate the functionality so that IT administrators have deeper levels of control and greater programming ability. With SDN, the administrator has virtualized components that makes it easy to automate the way virtual network connections can be set up to handle traffic flows (network service chaining) and also “spin up” networking environments without all the manual work. Plus, they can change any network rules–prioritizing, de-prioritizing or even blocking specific applications, traffic, and packets. These are the benefits we’re recognizing today, but fundamentally it will make much more possible. SDN lays the foundation for the networks of our future.
SDN’s decoupling aspects provide a layer of abstraction that positions IT experts to design even more forward-thinking capabilities based on control, automation, and intelligence. This abstraction evolves networking from “IP-address-aware” to “business-intent-aware.” It represents the essential building blocks that networks need to operate at a “business-intent layer,” giving IT administrators the ability to create self-provisioning networks. In fact, “business-intent networking” is the next step on the path to self-aware, self-adjusting, and ultimately self-driving networks.
Intent-based is the next generation of networking that will become self-aware, self-adjusting, and ultimately self-driving.
While our quick social poll demonstrated that most people are not familiar with the term intent-based networking, the majority of respondents (80%) believe we will have self-driving networks in the future. Let’s dive a little deeper into how intent-based networking works and how it fits into our evolution path.
Today, SDN allows programs to modify and configure the network dynamically, but it still requires someone to translate the enterprise business requirements into network design elements. In the future, however, we will give networks our high-level business needs, and it will understand that intent, monitor itself, and be able to modify the underlying transport infrastructure to tune the platform itself–all in real time.
The network of tomorrow will trade today’s “path-selection rules” for intelligent network configuration and real-time re-configuration. Instead of simply picking the most appropriate path for application traffic, networks will dynamically design the network to match business motives and manage traffic demands. We will stop commanding networks how to do the job and instead tell them what to achieve. To make this leap, networks will leverage AI technologies to be self-learning, so they can recognize patterns, predict usage, anticipate needs, and achieve your desired business intent. These distinct self-provisioning advancements will deliver self-sufficiency like we have never seen before. We will have auto-managed networks.
To understand these developments, let’s look at the maturation of network controls and configuration.
The evolution from programmable, SDN-powered networks to intelligent, self-adjusting networks is apparent when we look back at network development. In the early years, networks were governed by physical switches and transmission lines, which were essentially fixed cable analog systems with end-to-end equipment. Over time, IT developers abstracted the switches so they could easily manage them, hand crafting rules to command them.
Then, the advent of computing virtualization and resulting public cloud services required an automated approach. SDN was born. Its advancements decoupled the entire control plane, allowing programmable software, logic functions, and upstream systems to take control. SDN started in the data center and naturally expanded to WANs. More recently, end customers have been given access to that programmable software via online portals and mobile apps where they can manually and remotely modify the network or leverage their own software (via APIs) to make modifications. In fact, investment in these network automation API tools is expected to increase: By 2021, productized network automation (NA) tools will be utilized by 45% of organizations, up from less than 10% today. SOURCE: “Predicts 2018: Enterprise Network Infrastructure and Services” (G00344586)
In the next step of the evolution, machine learning and AI-driven innovation will replace humans in anticipating and predicting the changes necessary for network management. The computer will make adjustments in real time, without human intervention. The days of mapping applications to IP addresses and application performance requirements to classes of service will be relegated to the fate of the punch card. Not surprisingly, enterprise CIOs are expected to respond positively. According to Gartner, by 2020, more than 1,000 large enterprises will use intent-based networking systems in production, up from less than 15 today. SOURCE: “Predicts 2018: Enterprise Network Infrastructure and Services” (G00344586)
How do you create a self-modifying network?
First the network must have the ability to identify the behavior (see it in real-time), then learn the behavior, and then make changes to it based on the evaluation. These are the foundational stepping stones of a self-modifying network, and software defined technologies make it possible. Once you have each of these elements abstracted, networks will be able to evaluate the behavior, detect patterns, and programmatically create response plans and actions.
SDN puts the future within reach, and Masergy is positioned to seize the opportunity. Masergy has done more than stay in-step, it has been driving the leading edge in networking solutions.
Since its inception, Masergy has designed its platform around the principles of software defined networking to make it highly programmable and infinitely adaptive. Masergy’s network bends to the needs of customers both in its ability to be custom fit for each unique enterprise footprint, application landscape, and user group needs, and also in its ability to be highly responsive, scaling and changing alongside the evolution of the business.
Masergy has long been shaping its platform around the needs of end user customers. When customers asked for calendaring capabilities that would allow them to schedule extra bandwidth for data backups and tools to program their network to behave differently during business hours versus the weekend, Masergy delivered. Calendarization tools were introduced in 2008. When enterprises needed a performance guarantee, Masergy responded again. The Intelligent Service Control enables global bandwidth on demand. Customer needs continue to define Masergy’s future. Today, enterprises want tools to write directly to our Software Defined Platform for which we offer open APIs. In addition to API capabilities, we offer iPaaS tools to streamline SaaS application integration and business process automation. We’re will soon be helping our customers secure all their SaaS applications through CASB (cloud access security broker) technology partnerships. As our way of continuously delivering built-in value, we’re always working on incorporating more tools, and opening more APIs. Finally, machine learning is a deeply ingrained of our security analytics engine, and continues to propel our solution suite through automated, data-driven optimization.
The next challenge is creating a self-aware network that will be intent driven, and this is our vision. Our team is exploring these opportunities, laying the foundations and pushing our boundaries into this new frontier. Both our people and our platform stand ready to welcome it all and share its advantages with you.
Learn more about the future of networking technologies: The New Era of Software Defined IT: Where We’re Headed Next
65% of IT teams are already using AI, and learning from them is the best way to shape your strategy. Join us as we unpack the latest AIOps research.
Here’s how our combined company will accelerate digital transformation and unlock more value for clients, partners, and the industry at large.
Collaboration applications won’t perform without the right support systems. In fact, it’s what’s behind the technology curtain that matters most.