'Breather' could signal stagnation of giant federal cloud/data-center-consolidation project

Plan to cut 2,100 data centers to 800, cloudify the remainder, may disappear with its author

U.S. government agencies are breaking tradition to be among the first to adopt something new and exciting from the world of technology.

Large bureaucracies tasked with delivering services across the breadth of a large country, without the margin for error or ability to decide not to deliver the service after all – as commercial companies can do – don't leap at unproven processes or technology.

That's exactly what they're doing, though, including the chance they'll fall flat on their faces while continuing the 25-point "cloud-first" total-government migration plan created under the leadership of U.S. CIO Vivek Kundra, who is quitting that job for the second time.

Kundra introduced the plan in 2009, in conjunction with the Obama White House, and put it place in December as part of an overall "e-government" cost-saving efficiency raising effort that included regular progress reports to the public and struggles with budget cuts and other bureaucratic potholes.

Kundra, the former CIO of the District of Columbia was appointed to the post of U.S. CIO months after the Obama adminstration took office.

Among his highest-profile, highest-impact goals was to consolidate more than 2,100 data centers to 800 by 2015.

The other was to modernize and consolidate federal IT data centers using cloud technology to continue increasing their efficiency and cutting their costs.

Federal CIOs and analysts credit Kundra with energetic leadership and significant progress in an area that has seen little of either ever in its history.

Through debate and tooth gnashing at the closure of so many facilities, and objections to the adoption of technology that hasn't been proven in the private sector, Kundra kept the discussions, if not the project itself, on track.

He announced in June he would resign to take a fellowship at Harvard's Kennedy School of Government.

A third of the 25 milestones in the plan should have been accomplished by now, progress hasn't been nearly that fast, according to Computerworld.

The National Institute of Standards and Technology (NIST) did create a standard definition of what constitutes "cloud" computing, and an outline of how to make them secure, which must have helped.

A new survey shows that federal IT may not even know how many computers they have or what is being done with them, let alone how quickly they're consolidating all that hardware onto virtual servers and then into a cloud.

Of 157 agencies surveyed, only 25 percent said they're tracking the amount of storage they own or use and 41 percent said they know how many servers they operate.

Only 31 percent of federal data-center managers knew the average computing load across their data centers, compared with 90 percent in private-sector companies, the survey showed.

Despite sometimes-gloomy reports on the progress the feds are making toward cloud, it's possible the agencies have no idea how much progress they've actually made, or not.

Some of the remaining CIOs want to take a breath to make sure the cloud plan is in good shape. Many of the rest are optimistic about completing the plan – or at least continuing it.

"Now you have a perfect situation where the stars are aligned to make massive change," cloud fan Shawn Kingsberry, CIO at the federal Recovery Accountability and Transparency Board told Computerworld. "When you look at what this means, federal government has the opportunity to make moves forward."

Moving things forward is the thing the other CIOs praised in Kundra. Many described him as a "change agent."

The thing about change agents – at least in chemistry, if not in politics – is that when you remove them, the change stop.

In bureaucracies – government or commercial – the same thing happens if you pull out the money.

The Dept. of Interior just cut $500 million from its IT budget; other agencies may follow suit.

The decision to consolidate the number of federal data centers by 60 percent, layer cloud platforms over them to make better use of their resources and improve their ability to share data was uniquely bold and visionary, especially in the federal government and most especially among projects that actually got off the ground.

Without the catalyst credited with articulating planning, evangelizing and launching the project, the chance that it will continue on the same track is small. The chance that it will continue to a conclusion that delivers enough to declare victory even with a fraction of the progress in the plan is also small.

Far more likely, with parties in Congress refusing even to disagree with each other, let alone work out any differences, budgets shrinking, milestones being missed and the project leader on his way to academia, the odds are very strong that the federal cloud unity project will be declared complete months or years in the future without making a lick more progress than it made the day Kundra quit.

The frustrating thing is that it's just as likely that Kundra quit because the project had hit a permanent stall, not caused it by leaving.

Which would mean federal IT is even more hidebound and resistant to improvement than even the most pessimistic expectations.

From CIO: 8 Free Online Courses to Grow Your Tech Skills
Join the discussion
Be the first to comment on this article. Our Commenting Policies