BACKGROUND IMAGE: iSTOCK/GETTY IMAGES
BOSTON -- The rise of the digital transformation economy -- and the steps needed to enable that shift -- highlighted IDC's annual Directions conference, held here last week. While it is oftentimes easy to get caught up with emerging technologies, like artificial intelligence, the internet of things, and augmented and virtual reality, the underlying message is much clearer, analysts said: Without network transformation, fueled in large part by SDN and automation, there is no digital transformation.
"From my perspective, to achieve true digital transformation -- or let's call it digital enablement of the enterprise, which is a big theme -- my network needs to transform, too," said Rohit Mehra, an analyst at IDC. According to Mehra, a transformed network is an intelligent one that allows enterprises to not only have a robust pipe for applications and communications, but also the ability to create policies that yield strong security, high performance and low latency.
"Essentially, if I cannot make my network dynamic, I will fail in my digital-enablement effort," he said.
Unifying business goals with networking goals
A dynamic network also lets enterprises determine what they want to achieve with their operations, Mehra said. Instead of focusing on targeted IT products, enterprises are looking at their business goals and figuring out how they can align those goals with their networking objectives.
"Instead of looking at the core tenets of networking, we're now looking at what applications we have," he said. "We want to align those applications via network intelligence to the underlying network. That then makes my network secure; it makes my network dynamic."
That's where SDN architecture becomes a transformative agent.
"You can't buy SDN as a product. But architecturally, it is a key technology that will have broad-based deployment in the enterprise, as well as among cloud service providers and data centers," Mehra said. "[IDC] thinks the market for SDN spans all segments -- all kinds of service providers, whether it's telco or the cloud guys, but also the enterprise campus."
Automating the network for more efficiency
Deployment strategies will vary, with service providers coupling SDN with open source and white boxes, while enterprises still require a fully integrated, vendor-provided SDN option. Cloud providers have more flexibility to use white boxes because they have the capability to run whatever software component they want. Enterprises, Mehra said, don't have that ability. For that reason, he said, enterprises are still far away from adopting open source software and commoditized hardware on a widespread basis.
"Automation is one of the key goals that comes out of network transformation, and you cannot achieve automation without SDN," Mehra said. SDN also harnesses machine learning, a technology that underpins digital transformation. Case in point: An intelligent network discovers packets are being dropped, and before any human needs to intervene, the network self-corrects automatically. Thus, Mehra said, IT staff members can focus less time on "keeping the lights on" and work on other areas of the network.
What are some of the biggest challenges when using machine learning with SDN for network configuration management?
SDN analytics will make the network smarter
SDN and NFV key to network transformation