Slashdot Mirror
US Government Data Center Consolidation Behind Schedule, Cost Savings Uncertain
itwbennett writes "The goal of saving $3 billion by closing 1253 data centers is 'very realistic,' says David Powner, director of IT management issues at the U.S. Government Accountability Office — except that agencies haven't been able to track cost savings for the initiative. Eighteen months from the 2015 deadline, 'we have no idea how much we've saved the taxpayers,' said Steve O'Keeffe, founder of MeriTalk, an online community for government IT issues. This isn't the first snag in the project. Almost a year ago, Slashdot reported that the project was woefully behind schedule."
The government released a summary of what data they do have (PDF), and at least the DoD expects to save $575 million next fiscal year. Also see the full GAO report.
I'm guessing that they'd be able to track those metrics very well if they tried. My experience with the PHB management types is that they don't want to do anything unless they absolutely have to, especially if that something would create accountability for them or anyone on their teams.
Rather than do it themselves in less than two years they've hired "contractors" and they have no idea what the business case for this consolidation is because they haven't hired a consultant to do the evaluation yet.
The answer is probably "0"
The cow says "Moo." The dog says "Woof." The Timothy says "Thanks, valued customer. We appreciate your input."
Powner, Director of IT? Can't make that up.
Ever notice how easy it is to spend other peoples money? they have no real reason to try and save that 3 billion because its not from their pockets. just raise taxes again and it will all be ok. In the end this will probably be like everything else
"we know we told you this project would save you 3 billion but we made an error"
what was the error?
"well we meant to say cost, not save. this change will cost you 3 billion"
have you seen my sig? there are many others like it but none that are the same
If, as even the Democrats admit, the government is "too vast" for the President to be responsible, we ought to be shrinking it, rather than continue its expansion to the new domains (like, uh, healthcare)...
In Soviet Washington the swamp drains you.
How much have they saved now? Probably not much, and probably spent more.
It costs money to find the location, arrange the contracts, hire new people/lay off or move staff, and move equipment between rooms. Even then, many organizations may say it's cheaper to buy new hardware than it is to move 5+ year old servers.
However, the efficiency once many of those data centers is closed will become apparent. Space/power/cooling/networking/staff isn't cheap. Consolidating will give you some amazing savings a few years in.
Maybe we should slow down with the deficit reductions; this consolodation may not be necessary and may be detrimental.
You know you're in trouble when they speak of how much money they're saving instead of how much they spent to acquire a feature or benefit. What about the data center that stores every digital interaction in America? I bet that costs a pretty penny to operate, let's shut it down, it's not doing us any good. Didn't stop the Boston Bombing... Hell, zoning laws could have prevented the recent explosion in West, TX (which was more harmful than the Boston bombs), but I don't see them trying to save anything at all anywhere -- Not even lives.
"Honey, look what I got, you wouldn't believe how much money I saved today!" -- No. She spent money, didn't "save" a dime. I'll evaluate cost to benefit ratio to determine if the purchases were wise.
Want to save money? Why not get rid of the DHS? They're not needed. We have FBI and cops already. We don't need a huge cumbersome annoying workforce of security guards who don't actually provide security. You Can't Provide Security for others -- They can only protect themselves, and should be aware of surroundings and cautious of dangers if need be; That saves money and lives.
How much of the money they're spending on 3rd party contractors is wasted by inflating the costs to turn a profit? I'm not stupid. They haven't saved a dime one.
Government program behind schedule and over budget exactly according to plan.
Some of us saw this coming since the day it was announced.
a) This is big government we're talking about. We're lucky the costs aren't higher.
b) Since when has anything orchestrated by this administration been anything but the opposite of what claimed?
c) We're lucky they're actually telling us at this point.
~/ssh slashdot.org ssh: connect to host slashdot.org port 22: too many beers
I work for a large DoD organization in MD as a contractor. I have worked here for a 1 year. Prior to that most of career was in commercial software engineering world. The level of waste, irresponsibility, small "kingdoms", personal ownership of public domain and laziness is unimaginable. There are people who make 100K plus but are completely useless because they refuse to do any work, alarmist, or get in other peoples ways and prevent them from being productive.
The government is not "vast" or "large" rather it is "stupid", "childish" and "irresponsible". Specially DoD then again what do we expect from an organization that is primarly run by people who are too dumb find their own asses. These people enrolled in the DoD (Army, Navy, AF and others) to take order so they don't have be bothered with any decisions or actual responsiblity in life.
...fancy that.
"How do you think you'd feel if you knew you were on top of two million parts built by the lowest bidder in a government contract?"
They need another datacenter to hold the computers that track the cost savings from the other datacenters.
Hovernment does not know how to budget and Software is always behind schedule.
Put the two toigether and you'll not have any savings at all but a growing cost.
To get a project done, you have to want to get it done.
Server and data center consolidation projects require some interest on the part of eliminating or consolidating old systems.
Government workers, in my experience, have no interest in server consolidation.
If they built a system using a 286 processor, and have no choice, they'll keep using the 286 based system.
Tis unfortunate, but it's what I've seen.
Data center consolidations are never easy, and this is most likely being handled by a for-profit contractor, so tack on an extra 20% to whatever price tag it is, just for overhead.
Think of all the stuff that has to happen:
- All the connectivity to various networks has to be moved or duplicated. If we're not talking IPsec over the Internet, that means circuit orders, routing changes, etc. which quickly multiply and all involve tons of coordination.
- If you're doing P2V, that has to be carefully scheduled and everything needs to be tested on the consolidation hardware before you decommission the old stuff completely.
- If you're doing lift-and-move for things that can't or shouldn't be on VMs, downtime, network changes and logistics to get it to the new place need to be coordinated.
- All the system dependencies need to be worked out. System X will break without connectivity to Systems S, T and Z. System T will fail intermittently if the latency between System F goes above a certain point, taking down A, B and N. Etc. Etc. Yay for integration.
Also, DC migration plans tend to expose all the skeletons that previous admins left in the closet. Well-worn examples of undocumented networks, networks that unexpectedly rely on some crappy Linksys switch backboning traffic between key segments are par for the course. [1]
My experience from the private sector with consolidations is usually large companies wanting to shovel things that started off in branches up to a central point, or to move IT someplace with a lower cost of living. In every case, it's never paid off right away. The fact that it's the federal government just means they have more real estate at their disposal, so it's a bigger job, not necessarily more or less cost.
[1] My favorite comes from the dark ages of Token Ring where it turned out that the link between two components of a very important, must-be-running system depended on TR gateway software running on an original IBM PC AT with two TR NICs.- and this was in the 90s. It was never meant to be that way, but it turned out that someone never documented this and it was found during a move.
I am a DoD sysadmin.
The 1,200 data centers comment is misleading. The DoD classified a closet with a few servers as a "data center". Only a hand full of actual data centers exist. Eventually all those small closet "data centers". Will be gobbled up by the hand full of big ones. The cost cutting will be giving control to admins in the central locations. Closing the small ones and moving the servers to virtualization.
The big question is...who? GSA, DISA and private firms are all lobbying for control.
A few years ago, we got a "new" data center. I was all excited to move our old and busted applications at least off of old an busted hardware!
However what we got was a new building.
Rather than getting new servers, what we actually got was a few days of downtime while they unplugged and loaded it all on a truck and drove our busted hardware to the new building! Progress!
When I was told, I had one of those laughs... Although all I could think of was what would happen if the semi truck carrying all our stuff crashed... sure it is all backed up, but man we would have been screwed.
I work for the DoD (Army) and have seen the datacenter consolidation progress in my organization and others in my Theatre. There is no method or order to collect energy savings analysis, so we are never going to know how much money/power is being saved. No one, at least in my Signal command or lower, ever asked for it.
Of course, being that the consolidation is mostly-complete in my Theatre, its too late go gather any real data. Most likely, Brigades will simply make estimates based on the number of servers then and now.
We also have nearly every user workstation hibernating after ten minutes of user inactivity as part of the three billion dollar energy-saving presidential order. No data has been collected on how effective that is either.
Its all typical DoD kneejerk bullshit, no one knows the true impact, they just want to finish before their peer Commanders to look good for promotion.
It will cost millions without any gains in efficency. The new facility has to upgrade their utilities for us to move in. We are moving from a Government facility to a Leased one. Bring your own chair there is no room for you.