The enterprise is undergoing a process of digital transformation in which longstanding business models are being updated for the age of digital services. Naturally, this is having a big impact on data infrastructure, particularly the network.
But while change is good, it is not always easy. And when the change is rapid and far-reaching, as it is for the network now, it can be very easy to lose your way. So what, exactly, is “network transformation,” and how should the enterprise go about it?
According to Market Insight, network transformation is shaping up to be a $330 billion+ business by 2025, representing a whopping 62 percent compound annual growth rate. The field is varied, of course, and represents everything from new physical infrastructure to virtual or software-defined solutions in the data center, the cloud and on the edge. The goals are wide-ranging as well. New solutions are expected to reduce latency, increase flexibility and simplify management burdens, even as they lower costs and streamline the overall network footprint.
It’s fair to say that this movement has shifted into high gear, says Cisco’s Vivek Bhargave. The recent Cisco Live event was chock-full of new DNA Center partner solutions ranging from power consumption analytics and automated on-boarding tools to newly intelligent systems, services and devices. The key to all this activity is the open API framework that allows third-party systems to interact with the platform from all directions. Northbound APIs, for instance, abstract and automate manual tasks such as monitoring and performance management, while east-west APIs enhance core functions like IT service management and IP address management.
Artificial intelligence (AI) is also fueling this process, particularly the machine learning (ML) algorithms that are making connectivity to the Internet of Things (IoT) not only possible but optimal, says Bruce Davie, CTO for Asia-Pacific and Japan at VMware. This is actually a two-way street because the more that services related to connected devices come to rely on AI for their basic functionality, the more the networks need to transform in order to accommodate these new workloads.
How will they transform? In part by incorporating the same AI and ML tools within their management stacks. Going forward, a key challenge will be implementing a unified virtual cloud network that extends management and security frameworks across all layers of the network, from the data center to the cloud to the edge and beyond.
Perhaps the biggest challenge in all of this, however, is paying for it. But as ISG’s Dave Muller points out, the network itself may be able to help out through better funding mechanisms. Most organizations are already reducing their network budgets through strategic sourcing — some by as much as a third as the cost of services continues to decline. Meanwhile, the shift to a more operational footing on the capital side allows the enterprise to shed the high-cost practice of simply adding more capacity when loads start to push the envelope in favor of a more budget-friendly continual optimization scheme. The money saved, of course, not only fuels network transformation but overall digital transformation of the entire business model.
It’s hard to say what the network will look at after the transformation is complete. For sure, it will be more abstract, more automated and more dynamic. Beyond that, there are countless ways in which APIs, protocols, platforms and other technologies will meld to produce the highly connected world of the future.
And in all likelihood, there won’t be one approach to enterprise networking, because the ability to craft customized architectures will be so much easier in software.
Hang on, it’s going to be a bumpy, but interesting, ride.
Arthur Cole is a freelance journalist with more than 25 years’ experience covering enterprise IT, telecommunications and other high-tech industries.