Live data center modernization: A nail-biting experience

Modernizing a live data center with no downtime is no small task, but a data center colocation company managed the nail-biting task.

NASVHILLE, Tenn. -- Updating a data center is not for the faint of heart.

For more data center modernization and design:
Data center budgeting: How to convince management to upgrade.

Calculatingcosts of data center cabling infrastructure.

Todd Gale likened his company's recent data center modernization to picking up a running locomotive off one track and placing it on another without disruption.

The task was to upgrade a 25-year-old data center, built by AT&T during the 1980s to handle mainframe computing. The building was 125,000 square feet; the data center space, 80,000 square feet. In this case, modernization required ripping and replacing most of the facility's power and cooling infrastructure -- all without incurring system downtime.

Gale, the regional data center director for the Denver-based colocation company ViaWest, led the charge, and the process took 18 months.

"We had two major tenets," Gale said at the AFCOM Data Center World conference this week. "We wanted to keep the tenants informed. We have some large customers, some Fortune 50 companies. Also, we wanted 100% uptime. That was spelled out very clearly to me by the chief operating officer."

Getting to 100% uptime: No cutting corners
The facility's critical updates included replacing all cooling towers and one-third of the air handling units, upgrading the chiller plant, and increasing capacity on the cooling side. On the power side, the plan called for replacing a massive, outdated backup power system; expanding the power distribution panels; and installing power quality meters on the utility feeds. Because of the facility's size and lean staffing, Gale noted, the company also needed to install a comprehensive building management system.

 We wanted 100% uptime. That was spelled out very clearly to me.
Todd Gale,
regional data center directorViaWest

Fortunately, AT&T built N+1 redundancy into the facility, meaning Gale could take a cooling tower offline and two others could still run the data center.

"If anybody in your finance department says you don't need this extra stuff because it costs money, this is a perfect example of why you need it," he said.

When deciding on which air handlers to install, ViaWest conducted a bakeoff and, according to Gale, Stulz was the clear winner. Ditto on the uninterruptible power supply (UPS) side, where Mitsubishi got the nod.

The results were clear. ViaWest saw a 50% improvement in the operating cost of the plant, its power usage effectiveness (PUE) improved by decreasing from 2 to about 1.45, and the company garnered about $1 million in rebates from the local utility.

"You have to embrace any project like this," Stulz said. "It's nice to work at a company that comes to me and says you can build more stuff. They're exciting projects and it means the company is doing good."

Mark Fontecchio can be reached at mfontecchio@techtarget.com.

This Content Component encountered an error

Pro+

Features

Enjoy the benefits of Pro+ membership, learn more and join.

0 comments

Oldest 

Forgot Password?

No problem! Submit your e-mail address below. We'll send you an email containing your password.

Your password has been sent to:

-ADS BY GOOGLE

SearchWindowsServer

SearchEnterpriseLinux

SearchServerVirtualization

SearchCloudComputing

Close