The Defense Department came under scrutiny in March after its inspector general found the agency would fall short of its internal goals for data center consolidation and those mandated by White House policy.
The U.S. military operates thousands of energy-guzzling data centers on U.S. soil, overseas and even in the seas – the Navy’s ships and submarines are essentially floating and submersible data and computing centers. Despite criticism, its goal is to shift to more efficient, less energy-consuming computing platforms.
As DOD Chief Information Officer Terry Halvorsen has explained in the past, the logic in closing data centers is potential savings in real estate, power consumption and perhaps increased efficiencies and scalability through platforms like cloud computing, but the single largest driver is labor costs.
The labor issue can also be the biggest roadblock to actually closing data centers, as Gary Wang, the U.S. Army deputy CIO/G-6, explained Wednesday at an event hosted by FCW.
"The expectations when you move to cloud are really, really huge savings," Wang said, but in reality, full recouping of costs for buildings or facilities is rare as the government often continues to own and maintain those facilities anyway.
In terms of labor, while 100-percent contractor-operated facilities do exist, they’re also rare, which causes a conundrum for top brass when facilities targeted for closure are occupied by a large contingent of military or civilian federal employees.
“Generally, when you look at IT shops, it is 80/20 or 70/30 [contractors to feds],” Wang said. “So what happens to all the government billets is you have to find jobs for them, you don’t get to let them go. In industry, you can lop those guys off. In government, you don’t have that.”
While it’s no secret that laying off federal employees can be challenging and it generally isn’t the military’s approach to such matters, military agencies have used such approaches before. In 1994, the U.S. Navy used a reduction in force to abolish 700 occupied jobs in a Navy command that totaled 3,200.
Wang said handling the labor issue can be more complex than the technical hurdles inherent in moving from legacy IT environments to modern ones. It can also be more expensive.Wang said Moore’s Law continues to drive down the cost of hardware, but it isn’t reducing the cost of owning and operating facilities or employing unnecessary workers.