The main goal for any enterprise data center is to be “smart” — to predict failures, understand the physical environment, and recognize threats through automation and tools. With data center technologies such as intelligent power, sensors, artificial intelligence and dynamic cooling, such smart data centers are now a reality.
Smart data center technologies, and their role in everything from colocation facilities to internet of things (IoT), was a prevalent theme at IDC Directions 2017 in Boston last week. Enterprise IT shops should take care, though, not to rush to choose and implement them.
New colocation features
As the colocation market saturates, providers seek ways to differentiate themselves. One possibility is augmented reality, which has gained adoption in the medical field to provide additional information to surgeons mid-operation.
This technology isn’t a reality in the data center yet, but it could offer colocation customers a smarter alternative to a remote hands service, said Jennifer Cooke, a research director of data center trends and strategies at IDC in Framingham, Mass. For example, an IT admin witha pair of augmented reality glasses could remotely walk someone through steps to troubleshoot a server problem.
“Beyond saying, ‘This is your empty space where you can plug things in,’ [colo customers] are looking for more advanced services,” said Cooke. “As people come to grips with the fact that the data center is not their home data center but a collection of resources, that’s when the tools are especially valuable.”
IoT warrants remote tools
The advent of IoT brings the need to process massive amounts of data, but not necessarily at a central data center. By 2019, organizations will process 43% of their IoT data on hardware that rests on the edge of a network, IDC predicted.
“If you think of the data center itself, it’s almost like a little IoT ecosystem,” she said. “You’re generating a lot of data and trying to make sense of it — it’s kind of like a test bed for IoT readiness in a way.”
If you think of the data center itself, it’s almost like a little IoT ecosystem. Jennifer Cookeresearch director of data center trends and strategies at IDC
As more data is processed remotely, admins will need more remote control capabilities that smarter data centers offer, Cooke said.
DCIM tools such as Vertiv’s Trellis and Schneider Electric’s StruxureOn fill that need with mobile applications, which offer visibility and central management to remote data center managers.
Bridge the IT divide
As the adoption of converged infrastructure and software-defined technologies rises, many organizations fail to address the increased power needs that accompany the transition. In fact, over 83% of companies that deployed converged infrastructure required an upgrade of their power and cooling infrastructure and nearly 30% experienced delays in deployments due to those power and space constraints, IDC reported. Smarter data center technologies can help to bridge that gap between IT infrastructure and facilities.
Real-time analytics and remote visibility help data center operators plan their capacity requirements for an infrastructure overhaul. Unprepared IT organizations, she said, are like children who believe writing a check causes money to magically appear in a bank account.
“In a way, IT is like that kid: ‘Just plug it in, the utility will be there,'” she said’ “But they don’t consider that they might shut down the other half of the room, or there’s just not enough power to support [the transition to new infrastructure].”
Converged infrastructure can also throw off the cooling ecosystem because it will pump out a lot of heat in one area and have a denser footprint when it comes to power consumption, said Cooke. Cooling innovations such as in-the-door cooling directly targets the servers themselves, lessening the burden of the infrastructure on the rest of the data center.
To avoid the facilities and IT gap, Cooke predicted that IT pros will adopt better management technologies, or turn to colocation and service providers for a hands-off approach.
