CitiGroup LEEDS by example

Network World - GEORGETOWN, Texas -- Jim Carney, executive vice president of data center planning for Citigroup, likes to describe the company's newest compute facilities as "24/7 by forever."

Inside a green data center

In data center life, that "forever" translates to a good 20 to 30 years -- at least for this New York-based global financial giant.

"We've succeeded in developing a very flexible platform that can adapt easily and seamlessly to changes in customer technologies and allow for differences in heat densities, power consumption and physical layout without causing severe interruptions to their service requirements," Carney says.

Jack Glass, executive vice president of critical systems engineering, says he likes to tell Citi clients this about the company's new data centers: "We can do anything anywhere but we can't do anything everywhere."

If a Citi business unit wants blade servers here and a grid there, the company's modern data center infrastructure can handle each request, Glass says. "That's how flexible it is. We can allocate that capacity and have high-density areas and low-density areas -- our hands aren't tied in the data center space," he adds.

Global view

That wasn't always the case. Five years ago, Citi faced a number of challenges relative to its aging data center infrastructure. Capacity was exhausted, flexibility limited and "proximity risk" too high because, at the time, Citi had concentrated most of its processing in major metropolitan hubs such as London and New York.

Out of those challenges arose a cohesive, global data center strategy. "What we had been doing was incremental and now it was time to take a big step," Glass says.

The plan calls for Citi to reduce the number of data centers worldwide from 52 to 14 strategic sites, including five constructed from the ground up. The latest of its new North American data centers, in Georgetown, Texas, went online in January.

Citi, not unlike many companies, faces an interesting dilemma as it undertook this global data center strategy. While it certainly upgraded data centers here and there and expanded others, it hadn't built a completely new one in 20 or so years, Glass says.

But it quickly learned a few essentials, and the Georgetown data center represents the culmination of all the lessons, Carney says.

No. 1, the data center team realized it needed a global standard for data center construction and engineering. "We're not dictating to the Nth degree all of the components but putting in place criteria for reliability, maintainability and redundancy," he says.

And two, teamwork matters. Carney's data center team provided the engineering perspective and the corporate real-estate representatives lent the construction expertise to the projects, he says. "We brought together the same team throughout the strategy … so we could build upon each project and apply the best practices and lessons learned as we went along."

Reprinted with permission from

For more information about enterprise networking, go to NetworkWorld.com
Story copyright 2009 Network World, Inc. All rights reserved.