Army tackles data center consolidation, enterprise e-mail

Maj. Gen. Mark Bowman is director of architecture, operations, networks and space at the Army’s CIO office. He is responsible for establishing and implementing strategy, policy and governance for Army networks.

He spoke with Defense Systems Editor-in-Chief Barry Rosenberg about data center consolidation, the migration of Army e-mail to the DISA enterprise and the CIO’s strategic vision.

DS: What’s at the top of your to-do list — the thing that was there last week, is there this week and will be there again next week?

Bowman: I would pick three. Comms and IT operations in support of the warfighter is always at the top of my list, no matter what day, no matter what week. Then I’ve got two big projects that have significant interest across the Army and across DOD. The first one is Army data center consolidation. We’ve got some folks that are working that one hard. We plan on closing 185 data centers by end of fiscal year 2015.

And the next one is enterprise e-mail implementation, and we’re doing that transition right now. We’ve got over 75,000 users on enterprise e-mail now across our Army, and we continue to migrate people every day.

DS: Migration to the DISA enterprise system.

Bowman: Roger.

DS: Let’s talk about data center consolidation first. Where does that process stand today?

Bowman: What we’re doing is taking advantage of base realignment and closure across the Army. That gives us a wonderful opportunity to see what should close and what shouldn’t close. In fiscal year 2011, we have intentions to close 27 data centers. Some of those are BRAC locations, and some are not. Some are guard activities, some are active-duty activities.

We’re very comfortable about this approach because we’re benchmarking off of what DISA has done in the past. And throughout the 1990s and into the early 2000s, DISA was getting a series of Defense Management Review Decisions, and they were told that they needed to have certain economies and consolidations that they had to achieve. So DISA went from over 190 down to 18 Defense Enterprise Computing Centers.

In our data center consolidation, we want to do this in a smart way with virtualization and start getting much better at server utilization. Across our Army, I would guess — I don’t have the statistics on this — but I would guess that our average server utilization rate is pretty low. It’s probably in the 30 percent range.

And that’s for a number of reasons. It’s inefficiency…somebody comes up with a new system and the first thing they do is get their own server so they can watch their data. With our approach we’re going to go to DISA first. We will go out to commercial activities and host our data out there when it meets mission needs and it’s the right thing to do. And last, and only last, will we have Army-only data centers.

DS: Is the consolidation effort more than a cost-savings measure? For example, will it improve security?

Bowman: It absolutely is. It’s a cost-savings effort, and we can measure that every day by number of facilities that we’re using, by electricity that we’re using, by [the] number of people that we have working on this. But it’s also security. So when we put all these things, and we have common baselines of security that these are all following our same top-level security architecture. It just makes a lot of sense, and we know what we’ve got out there. So it helps us in the security way as well, and that’s something that’s hard to measure.

DS: All the other military services have their eyes on the Army's e-mail migration to DISA. How do you feel it’s going?

Bowman: I’m actually happy with the way it’s going. I had an easy transition to enterprise e-mail and had wonderful service right from the start. They transferred my mailbox and it worked fine, but others have not had that kind of luck. We have put in [tactics, techniques, procedures] where you have to create a number of .pst files to put your data in and get the size of your mailbox smaller so it takes a lot less to migrate it. Our migration tool didn’t work as well as it should have in the beginning. We paid a lot of attention to that.

DS: Is that one of the reasons why I’m hearing people say that the per-seat cost is more than expected because that migration didn’t go as smoothly as expected?

Bowman: No, I don’t know who the "they" is that’s talking about the perceived cost going up. We don’t see that. That’s by agreement between Army and DISA, and I don’t see the per-seat cost going up. I know people are still out there saying, "I don’t think this is the right thing to do."

DS: Yes, lots of people are saying that.

Bowman: I was responsible for one part of DISA’s e-mail system as the chief of staff for the headquarters and a couple of the out-sites in the National Capital region. Alfred Rivera ran it for the rest of DISA around the world, and the performance that he demonstrated far exceeded what my guys were able to demonstrate; that’s just a fact.

So I actually did have confidence in DISA doing this, and I still have confidence. What we’ve learned is it is clearly the right thing to do. Migration is at least as tough as we thought it was. Different places have different levels of configuration management, so when people have different variations, or not-so-good record-keeping or not up-to-date software, then it makes things a little harder than it should be.

So those are the issues that we run into. We have delayed migrating the Department of the Army Staff. The only staff section that’s totally done is CIO/G6. We’re on it. I hear guys complaining, “Oh, enterprise e-mail is down.” Well, in fact, you have to pull the string on it every time because it may not be down. It may be some other connection that’s down. It might not be the enterprise e-mail.

So a lot of people will throw rocks at it. Sometimes it’s warranted; sometimes it’s not. But for me, personally, the service I get is at least as good as it was before.

DS: What’s the road map for the rest of the e-mail migration?

Bowman: Our objective is to migrate the entire Army by the end of 2011. Now, [in June], we slowed down a little bit on our migration. We were going to do over 100,000. We scaled it back with some operational pauses in some areas, and we’re only going to do about 45,000 users this month. And we have really stepped that up. Initially, one of our goals was to be able to migrate 1,000 soldiers in a night. We can do that; we’ve demonstrated that.

DS: Is it a matter of a soldier leaving his desk on Tuesday night and having the migration completely transparent when he comes in Wednesday morning?

Bowman: That was it for me, other than I had a choice early in the log-on procedure. I could select either my legacy e-mail or my enterprise e-mail. The default is enterprise, and it’s been at least two months since I’ve been in legacy.

DS: What’s new and different in Army CIO Lt. Gen. Susan Lawrence’s strategic vision that’s different from her predecessor, Lt. Gen. Jeffrey Sorenson?

Bowman: OK, here is a curveball back at you. Gen. Lawrence is absolutely not new to the organization. She commanded [the Network Enterprise Technology Command], which was a direct reporting unit under CIO/G6 prior to this assignment, so she is still very operationally focused as one would expect. If there’s something out there that is slightly related to comms that we might be able to help with, we’re going to get involved. So our aperture is open a little bit. It’s not to say that Gen. Sorenson had it restricted or anything. She’s just stretching us out and pushing us.

She’s continuing actions that were started by Gen. Sorenson. One change I’ll mention is [the Global Network Enterprise Construct]. We want to move away from the term.

DS: Oh, really? You don’t like it anymore?

Bowman: It’s not that we don’t like it. It’s a construct; it’s a plan. Now, it’s all about building this global network enterprise, not focusing on some construct, some plan of the past. It’s putting the enterprise out there. The network ultimately enables all mission accomplishment by providing the capabilities and the services that are needed all the way to the technical edge in support of the warfighters.

Now to better accomplish this objective, we really need to take an enterprise approach across the Army to building out the network for our soldiers’ needs both today and tomorrow. Many organizations talk enterprise, many people get it, but when they think enterprise, they think within their particular functional area. The enterprise is all functional areas across our Army. So what we want to do is get to a true enterprise and get away from the constellation of functional enterprises that exist today.

DS: And they are still stovepipes, even though they’re called enterprises. Is that a correct characterization?

Bowman: Yes. So the global enterprise network will provide each soldier with universal access to his or her applications, data, critical ISR and video feeds, command and control information, continuous position location, mission update, collaboration tools and training capabilities. They’re all phases of our Army force generating cycle that we call ARFORGEN.

DS: What’s the enabler that’s going to allow you to go from those stovepiped enterprises to the larger enterprise?

Bowman: A common operating environment is key and critical to this. To-be architecture is key and critical to this. One of Gen. Sorenson’s legacies was the definition of what we have today and the to-be architecture. The to-be architecture is very simple. It’s everything over IP. So the way [of] the future is clearly everything over IP.

We have another fundamental driver in this, and we all know that the budget is going down. We don’t have the luxury, and we won’t have the money to be inefficient and have everybody develop their own [system]. So they’re going to be physically forced to do this, and they’re going to get operationally forced to do this because users need to be able to share data in today’s environment.