Having acknowledged in congressional testimony a few weeks ago that the Defense Department is not doing enough as part of the now 5-year-old Federal Data Center Consolidation Initiative, Terry Halvorsen, the department’s chief information officer, now plans to be much more “prescriptive” about what each military service and DoD component must do to rein in their costs.
Speaking to reporters on a conference call Friday afternoon, Halvorsen wasn’t ready to say what exactly he’ll prescribe, but said the department nowadays is less concerned about the number of data centers it owns than how much they cost to operate. And the cost of labor is the main metric DoD will use to decide whether a particular data facility is unduly expensive.
“The biggest cost driver we have in our data centers is people, and what we’ll look at is how many people are doing what. If you still have a lot of people doing monitoring and maintenance of servers, that’s a bad use of people,“ he said. “That can be automated today, and in our efficient data centers, that’s exactly what we’re doing.”
The DoD CIO’s office has launched a survey of all of the department’s data centers to determine their costs, with labor at the forefront. Halvorsen said once he receives the breakdown of highest-cost sites, he’ll issue new policies telling the Defense components which own them to reduce their expenses.
But if high labor costs are the main target of the latest initiative, it’s worth wondering whether DoD’s largest data centers might be targeted just for being large and employing a lot of people, and not necessarily for being inefficient. Some of those sites have already absorbed thousands of applications during previous DoD data center consolidation rounds.
Halvorsen said DoD would take steps to make sure it’s not cracking down on already-efficient data centers whose high personnel costs are just a function of their size.
“I’d like that to be my problem, but it isn’t. I’ve seen no examples of that.” he said. “But we will do cross-checking to make sure that we’re not inadvertently catching data centers that just happen to be big. We’ll look at labor factors, what kind of labor is being done, and I think we’ll continue to find out that [inefficient labor] is our single biggest factor. People talk about power, about location, and those are factors. But labor dominates our costs so much that this is really what we have to understand.”