By Steve O'Keeffe - http://bit.ly/ncBsLK
Two blogs this week. Focus on Data Center Consolidation savings. You can read the consolidated version now, and the built out version later - if it takes your fancy.
Snapshot - Come on Down
OMB published agencies’ data center consolidation plans on October 7 - all but DoD, who’s AWOL. More interesting than the plans - the savings numbers. Vivek Kundra championed FDCCI as the treasure chest to “pimp” Uncle Sam’s IT ride. So, how much will we save in 2011? Drum roll. $14.613 million. That’s 0.183 percent of the Fed IT budget. If that’s all you have time for, that’s the net up front. Thanks for playing FDCCI, the price is right. But, if you have time, think you’ll find it worthwhile to read more - to get the bigger story.
Big Picture - Bonus Round
Okay, here’s installment two, for those of you that would like to go to the bonus round. Before commenting on this savings number, let me go back to open government - since my last blog on the subject fueled such a reaction. Yes, OMB posted the FDCCI plans on time. But, wouldn’t it be a good idea to post a chart with agencies’ net savings up front as well as a progress on closing data centers?
I arrived at the savings figure by opening each agency’s report and hunting through for section six, where the agencies break out progress and savings. Or more correctly, some agencies break out their progress and savings. Many agencies did not even include section six - or their information was unintelligible. Might I be so bold as to call for a “Consolidated Consolidation Report?” It took some hours to piece together the data. People visiting OMB’s site could be forgiven for thinking OMB is not working hard to showcase the largess realized from FDCCI…
Here’s a rundown of who did what from my review of the reports.
How the savings break down:
- Agriculture $4.2M
- EPA $1.5M
- Justice $1.92M
- NASA $0.337M
- State $1.656M
- Transportation $5M
Total: $14.613 million
A series of agencies reported progress, but no savings:
Still others failed to report progress or savings – many missing section six altogether or their responses were unintelligible to me:
- GSA – included progress and savings, but completely nonsensical
- VA – included a lot of interesting futures
Some were required to report, but do not have any data center consolidations planned for years to come:
And, last but certainly not least, one department failed to report for duty, posting a note that it would fall in, in 30 days:
Now, you may be asking why did Vivek Kundra make so much noise about FDCCI? Wonder what the Wall Street Journal would make of these numbers after expending salacious ink - and the number was only 137 back then? Wouldn’t one expect bigger savings given OMB’s announcement that we plan to close 373 data centers by the end of 2012?
Let’s play with the numbers associated with these shuttered or condemned data centers. 373 is approximately 20 percent of the total population of Federal data centers - OMB pegged that number at 2,094. Data centers are expensive. Industry estimates put data center costs at between 10 and 50 percent of Uncle Sam’s total IT spend. Okay, I’m not great at mathematics - so let’s go with 50 percent, it makes the sums easier. If total Fed IT spend is $80 billion, and data centers are half the IT spend, that’s $40 billion. If we’ve closed 20 percent of the data centers, then we should save $8 billion. That number seems huge next to the $14 million in savings we realized from FDCCI in 2011. Even if data centers are just 10 percent of Uncle Sam’s IT purse, closing 20 percent of data centers points to $2 billion in savings.
To be sure, the financial savings are underwhelming and the quality of the reporting is not exactly first rate. There is some very interesting information in the plans - and they bear closer study, if you have the time to dig in - put on a pot of coffee.
The raw truth here is that data center consolidation is not easy. It’s not a quick fix. And, it’s not cheap. The recipe for success is complex. And agencies’ progress against super aggressive goals with zero funding to meet the mandate is precisely what one would expect. Poor.
This excerpt from the Department of Transportation plan says it all:
“It is a long process to identify possible consolidations, present them to management and then to users, and then work the technical side of migration. DOT has found that fast tracking these consolidations can lead to confusion from affected management and users and not enough time for technical issues to be resolved. Strategic communications is a very important activity to help gain stakeholders and to ensure users are aware of the technical activities that will affect them. One major activity that affects the schedule of consolidations is application mapping. Application mapping is a very difficult and time consuming activity, but it is also a step that cannot be skipped for successful completion of migrations. DOT has learned to concentrate on this step and validate data collected more than once.
“Migration costs have been more than was first reported. The complexity of the technical migration issues have caused schedule delays and increased costs in some migrations.”
Data center leads across government are struggling with this unfunded mandate. They’re having challenges putting together realistic plans. OMB’s goals and timelines are unrealistic. That’s the feedback from the Federal data center leads participating in the MeriTalk Data Center Exchange. Is data center consolidation a good idea - of course it is. But, we will not realize savings without the required strategic investments. It’s not necessarily about reducing the cost of IT, but rather about leveraging IT to increase the efficiency of government. Now’s the time for OMB to work with the agencies to build a realistic Data Center Efficiency plan to deliver real results on government’s bottom line. As GAO is working on an audit of the 25-Point Plan, perhaps now is the time for OMB to stop counting data centers and focus on more meaningful efficiency metrics - deal or no deal?
Interested in more data on FDCCI? Check out these research studies: