Americas

  • United States

The coming together of SD-WAN and AIOps

Tip
Mar 25, 20206 mins
NetworkingSD-WAN

SD-WAN delivers cost and resiliency benefits. Infusing AI into SD-WAN takes things further, enabling automated operations and business agility.

ai artificial intelligence ml machine learning vector
Credit: kohb / Getty Images

Software-defined wide-area networking (SD-WAN) and AIOps are both red-hot technologies. SD-WANs increase application availability, reduce costs and in some cases improve performance. AIOps infuses machine learning into IT operations to increase the level of automation. This reduces errors and enables businesses to make changes at digital speeds. Most think of these as separate technologies, but the two are on a collision course and will give rise to what I’m calling the AI-driven WAN

SD-WAN not a panacea for all network woes

SD-WAN is the biggest leap forward in networking since… well, the actual WAN. But many solutions still rely on manual configurations. SD-WANs certainly increase application resiliency, lower telecommunications costs, and often increase application performance, but they are more complicated than traditional WANs. Initial setup can be a challenge, but the bigger issue is ongoing operations. Manually tweaking and tuning the network to adapt to business changes can be time consuming and error-prone. A solution is needed to bring better automation to SD-WANs.

Enter the AI-driven WAN. Much like a self-driving car, an AI-driven WAN can make decisions based on different rules and adapt to changes faster than people can. Self-driving cars continually monitor road conditions, speed limits and other factors to determine what changes to make. Similarly, a self-driving network can monitor, correct, defend, and analyze with minimal to no human involvement. This is done through automation capabilities powered by AI, obviating the need for people to get involved.

Make no mistake, manual operations will hold businesses back from reaching their full potential. An interesting data point from my research is that it takes enterprises an average of four months to make changes across a network. That’s because maintaining legacy networks and fixing glitches takes too much time. One ZK Research study found 30% of engineers spend at least one day a week doing nothing but troubleshooting problems. SD-WANs can improve these metrics, but there’s still a heavy people burden.

With growing data challenges businesses face as they migrate to the cloud, they simply can’t afford to wait that long. Instead of being afraid of AI taking over jobs, businesses should embrace it. AI can remove human error—which is the largest cause of unplanned network downtime—and help businesses focus on higher-level tasks instead.

AI-driven WAN will transform network operations

So how will the evolution of SD-WAN into AI-driven WAN transform network management and operations? Administrators can use their time to focus on strategic initiatives instead of fixing problems. Another data point from ZK Research is that 90% of the time taken to fix a problem is spent identifying the source. Now that applications reside in the cloud and run on mobile devices, identifying the source of a problem has gotten harder. AI-driven WANs have the ability to spot even the smallest anomaly, even if it hasn’t yet begun to impact business.

SD-WANs are fundamentally designed so that all routing rules are managed centrally by administrators and can be transmitted across a network. AI-driven WAN takes it a step further and enables administrators to anticipate problems before they happen through fault prediction. It may even adjust network glitches on its own before users are affected, thus improving network performance.

A self-driving car knows the rules of the road—where the blind spots are, how to synch with traffic signals, and which safety measures to take—using AI software, real-time data from IoT sensors, cameras, and much more. Similarly, a self-driving network knows the higher-level rules and can prevent administrators from making mistakes, such as allowing applications in countries where certain actions are banned. 

Security is another concern. Everything from mobile devices to Internet of Things (IoT) to cloud computing is creating multiple new entry points and shifting resources to the network edge. This puts businesses at a security risk, as they struggle to respond to changes quickly.

Businesses can miss security gaps created by users, with hundreds of software-as-a-service (SaaS) apps being used at the same time without IT’s knowledge. Older networking technologies cannot support SaaS and cloud services, while SD-WANs can. But simply deploying an SD-WAN is not enough to protect a network. Security shouldn’t be an afterthought in an SD-WAN deployment, but part of it from the get-go.

Increasingly, vendors are bundling AI-based analytics with SD-WAN solutions to boost network security. Such solutions use AI to analyze how certain events impact the network, application performance, and security. Then, they create intelligent recommendations for any network changes, such as unauthorized use of SaaS apps.

Going back to the autonomous car analogy, AI-driven WANs are designed to keep roads clear and accident-free. They enable smarter networks that can adapt quickly to changing conditions and self-heal if necessary. With the growing demands of cloud computing and SaaS apps, intelligent networks are the future and forward-thinking businesses are already in the driver’s seat.

AI-driven WAN exists now and will explode in the future

AI-driven WAN may seem futuristic, but there are a number of vendors that are delivering it or in the process of bringing solutions to market. Managed service provider Masergy, for example, recently introduced AIOps for SD-WAN to deliver autonomous networking and has the most complete offering.

Open Systems, another managed service provider, snapped up cloud-based Sqooba to add AIOps to its strong network and security services. Keeping with the M&A theme, VMware recently acquired AIOps vendor Nyansa and rolled it into its VeloCloud SD-WAN group. That move gives VMware similar capabilities to Aruba Networks, which initially applied AI to WiFi troubleshooting but is now bringing it to its SD-Branch offering. Cisco is another networking vendor with an AIOps story, although it’s trying to apply it network-wide, not just with the WAN. 

Over time, I expect every SD-WAN or SASE vendor to bring AIOps into the fold, shifting the focus away from connectivity to automated operations.