On June 10, 2010, President Obama issued a Presidential Memorandum directing Federal agencies to sell off unused real property (land, buildings, etc.) in order to save the U.S. government some money. As the nation's largest landowner and energy user, the U.S. government was ripe for 'consolidation' opportunities. It made a great press release, but like almost all ideas borne of good intentions, things have gotten complicated from there.
First, since we are talking about real property, it seems it should be easy to find and dispose of. Well, not so much. The first problem is to know what the government physically owns. One would think that knowing what real property one owns would be a simple matter. After all, land doesn't move. With the exception of geologic processes, land doesn't change, and real estate is called 'real' because it's easy to see and touch. Besides, the government should have a lot experience with documenting its resources. George Washington was a surveyor; Mason and Dixon worked at government expense; and getting a handle on the nation's real estate was Lewis and Clark's raison d'etre.
But after 234 years since the Declaration of Independence and massive geographic expansion from the original 13 colonies, noise has crept into the system, and no one can say for sure how many acres the Federal government owns, where they're located, or in many cases, who has title to them.
Then comes the next problem. Once you've found a piece of property, how do you decide whether to keep it, sell it or consolidate it with something else? This is particularly vexing for properties that might host a facility of some kind that does not use surrounding real estate to its fullest extent - for example, something like a former naval air station that still hosts a coast guard search and rescue squadron. If you sell off the air station to a waterfront condo developer, where does the helicopter squadron go?
And then there's Congress, always looking out for its constituents. Senators and representatives are ever alert to any move that might curtail the flow of money, jobs or prestige into their districts. On Capitol Hill, pork is defined as money and resources flowing to someone else's district or state.
Data Center Consolidation Parallels
The point of this article is not to bash the Federal government but to point out the parallels between the government's real property consolidation drive with IT resource consolidation projects.
As with the Federal government, the first problem facing IT is for organizations to know what they have. Again, this sounds simple in theory but disconcerting in practice. Computers should be smart enough to identify and inventory themselves, right? Experience tells me no. In almost every IT asset inventory project I have ever been involved with, IT organizations discover that they underestimated the computer population by at least 30 percent, and that these previously undocumented assets almost always reveal themselves as biker bars of rogue devices, playgrounds for unlicensed software, and rotten boroughs of insecure and compromised computers.
I think this dates back to the early days of computing when Harry Truman's science advisor, Vannevar Bush, predicted that someday every country would have a computer. While intelligent robots were a staple of pulp science fiction of the era, it didn't really register with people that if you could stuff sufficient computing power to enable sentient artificial intelligence in an anthropomorphic mechanical assemblage, it would require then unimaginable miniaturization of computing hardware and that computers would be everywhere, not just one big one carefully installed in a National Hall of Science.
Once you get a handle on an IT asset base, it's time to make decisions. What do you consolidate, and where do you consolidate it? What can you salvage? What should be replaced? Can you make this switch at low cost with acceptable service interruption risks? And will the move really save money and be in your interests?
One can argue that any decision with regard to the above questions is better than no answer at all, but guessing wrong is rarely good for CIO promotion and career prospects. There is a big difference between raw asset inventory data and actionable information to guide and illuminate decision-making.
And then there are constituent relations issues that generally run along the lines of 'I support consolidation, but I really need my own X to do Y,' Or, in other words, 'Not With My Data Center, You Don't.' Some parochial concerns are in fact justified. Some computers really do need to run 24 x 7 with sufficient capacity (or even redundancy) to handle unpredictable load surges. Other misgivings are pure politics. Nevertheless, the best way to prevail in a skirmish or a life and death struggle is to marshal information superior to that deployed by a potential adversary.
Grit, Determination, and Consolidation
Consolidation isn't easy, whether it's reducing the Federal government's land holdings by a few hundred square miles or taking advantage of Moore's Law by moving workloads to more powerful and concentrated computing resources. But Federal and IT consolidations processes are similar. Know what you have. Identify what can be consolidated and where. Either make nice to affected parties, or bring it off before they can notice. And never underestimate the importance of perseverance.