You are here

Beating the Numbers Game

Data center consolidation is pivoting back to its original purpose: better data management.

Numbers. Some say they never lie. People love numbers and believe what numbers tell them, which is how numbers came to muddy the waters of the Federal Data Center Consolidation Initiative.

When then-federal CIO Vivek Kundra launched FDCCI in 2010, he noted that there were more than 1,100 data centers across government. He instructed agencies to inventory their data centers and begin the process of eliminating redundancy and waste.

The problem was — as he and others soon discovered — there were many more than 1,100 data centers, depending on how one defined a data center. And that’s where this worthy initiative about savings and optimization turned into something of a numbers game.


The percentage of agencies’ IT budgets that could be saved if they fully implemented five IT solutions: consolidation, virtualization, cloud computing, remote access and infrastructure diversification

SOURCE: ““Show Me the Money: The Key to Doubling Agency Savings” (MeriTalk, April 2014)

The actual number of federal data centers peaked at about 7,000 and led some people to question the IT community for being poor stewards of the public’s tax dollars. Cooler heads, however, such as acting Defense Department CIO Terry Halvorsen, cautioned against shuttering data centers just to force down the numbers. Yes, trimming federal data centers by 40 percent by 2015 and realizing $5 billion in savings are worthy goals. But to miss the forest for the trees would be to ignore a significant opportunity.

Forced to Optimize

In speeches, Halvorsen has suggested, in essence, that agencies simply cut money from their data center costs and focus on how best to optimize what is left. If that means closing data centers, so be it. If it means adopting technologies such as virtualization and cloud computing, rather than closing physical data centers (in buildings agencies have to maintain anyway), then pursue that path. At the end of the day, the goal should be to better manage the data, not the data center.

The Department of Energy might say, as Lawrence Berkeley National Laboratory’s Bill Tschudi told attendees of the GITEC Summit in May, that FDCCI presents an opportunity to make data centers drastically more energy-efficient. To that end, the DOE has published FDCCI guidelines specific to energy and water usage.

This brings us full circle. If you read the original Office of Management and Budget memo about FDCCI from February 2010, you will surely notice the mention of 1,100 data centers. But take note, in the same paragraph, of Kundra’s emphasis on energy efficiency. At the time, citing an Environmental Protection Agency report, he warned agencies that if they didn’t change their approach to data center deployment, energy consumption could double.

When he described the focus of the initiative in his 2010 memo, Kundra’s first bullet point was to promote sustainable IT practices. Further, he encouraged efficient computing, optimization and, yes, cutting costs.

“The cost of operating a single data center is significant, from hardware and software costs to real estate and cooling costs,” Kundra wrote.

There will always be those who want to know how many data centers have been crossed off the list (about 750 as of the last report, according to, but OMB was right, in 2013, to make FDCCI part of PortfolioStat and to shift the dialog back to optimizing IT resources. That’s really what FDCCI was meant for in the first place — to “identify potential areas for consolidation, areas where optimization through server virtualization or cloud computing alternatives may be used.”

The numbers are the numbers. And none of this is to say there aren’t wasteful, redundant data centers that should be mothballed. But there needs to be an innovative, deliberate, continuous effort to do things better. When agencies keep their focus there, the numbers will follow.

Tomasz Wyszołmirski/thinkstock
Aug 01 2014