Hyperconverged Systems Show Promise in Reducing Data Center Complexity

By Jeff Burt  |  Posted 2016-06-29 Print this article Print
Hyperconverged systems

Nutanix's Ting said it's the software that pools the nodes of a cluster, brings resilience and intelligence to the solutions and that drives the SDS capabilities critical for hyperconverged infrastructures.

"All that magic … is in the software stack," he said.

Nutanix in recent weeks has been pushing to broaden its reach in the industry. In May, the company unveiled its Xpress management software, which is designed to bring the capabilities of its Enterprise Cloud software to a price point for SMBs. A month later, Nutanix officials introduced a new offering that they said will transform the company from one that sells software into the hyperconverged space to one that provides a platform for the entire data center.

Software is important, Lenovo's Krishnan said, but the hardware vendor also plays a key role in developing the platform that the hardware runs on and offering services and support for those products. That is why partnerships like the one between Lenovo and Nutanix is important.

"This is complex technology," she said. "This is not something you can do in just six months. … You don't want to buy a white box and just put Nutanix on top of it."

Longoria said that despite the fast growth in the market, hyperconverged infrastructure vendors face challenges. One will be differentiating their products from those of their competitors. The hardware is similar and with Nutanix and VMware working with multiple OEMs, many are carrying the same software.

System vendors will be able to sell into their own customer base, but "for a new customer OEM acquisition, it becomes very hard to determine which software stack and which hardware partner you want to go with," she said. "To me, there's not a lot of good answers yet. Right now, I see everything being a sort of me-too offering."

Road to Composable Infrastructure?

Going forward, there also are questions about whether hyperconverged systems are a destination or a step along the path to composable infrastructure, where data center resources—compute, storage and network fabric—are pooled and treated like services. Organizations no longer have to configure hardware to run particular applications. Instead, it's the applications—using policies and APIs—that automatically and dynamically provision and deploy the infrastructure resources they need.

Some vendors already are moving into the composable space. For example, Cisco offers the UCS M-Series modular servers and UCS C3260 rack server, both of which company officials said are based on a disaggregated system architecture handled by the company's UCS Manager software.

For its part, HPE in December 2015 unveiled its Synergy platform, an architecture that ensures the exact amount of data center resources can be rapidly pulled together from a single resource pool to support an application and then returned to the pool when they're no longer needed. Synergy is being used by about 100 pre-release customers now, but will become generally available later this year, according to Paul Miller, vice president of marketing at HPE.

For smaller companies and midmarket customers, hyperconverged infrastructures will address most of their needs. "For enterprises, it's a stepping stone to composable," Miller said.

Both IDC's Sheppard and Moor Insights' Longoria see a similar pattern, with enterprises making the move to composable infrastructure ahead of their smaller counterparts.

"Hyperconverged has interesting capabilities right now, where composable infrastructure is not really mainstream right now," Longoria said. "But I think maybe OEMs see [hyperconverged] as a step to composable infrastructure. … There will be an interesting market transition point in the next two to four years."

BluePearl Makes the Move

For now at least, BluePearl Veterinary Partners is happy with the hyperconverged infrastructure technology supplied by Cisco and its HyperFlex offering. According to DePasture, the company's senior network engineer, the company has deployed three HyperFlex systems—including the first one to roll off the Cisco production line—and is awaiting a fourth to be plugged into the environment.

Company officials wanted an infrastructure that would enable BluePearl to run mission-critical applications and move patient information—such as diagnostic imaging—to any of their locations and that could scale. The result has been more effective and standardized veterinary care and an infrastructure that can support 100 percent annual growth at the company. BluePearl also has been able to reduce deployment and operational costs.

The move from VRTX to HyperFlex was a good decision, DePasture said.

"HyperFlex was comparable [in cost], but with everything that comes with it, we decided to make the change," he said.


Submit a Comment

Loading Comments...
Manage your Newsletters: Login   Register My Newsletters

Rocket Fuel