Slashdot Mirror
How the FCC CIO Plans To Modernize 207 Legacy IT Systems
Lemeowski writes in with this interview of FCC CIO David Bray. "When David Bray took over as CIO of the FCC last year, he found the agency saddled with 207 legacy systems, which is about one system for every eight employees in the 1,750-person agency. Bray, who is one of the youngest CIOs across the federal government, shares his plan for updating those systems to a cloud-based, common data platform, that's "ideally open source." In this interview, Bray shares the challenges the FCC faces as it upgrades its systems, including keeping up morale and finding a way to fit longtime employees into his modernization strategy."
Nice idea... I wish he could teach it to some of the politicians up there.
Good for him, he hit all the buzzword checkboxes. K street will have a lobbying job lined up for him when he's ready to golden parachute out of there.
Good luck there! How about offering them a "modern" early retirement/buyout package.
We tried that in the 1800's. It didn't work. For all of its flaws the civil service program works better.
putting the 'B' in LGBTQ+
I'm not talking about the janitors. I'm talking about pencil pushers who carry far too much influence. Their entrenchment is the only thing with the power to bring down the empire. The role of the elected official has become entirely ceremonial.
Maybe, but the legacy systems administrators are just high-tech janitors.
W..w..W - Willy Waterloo washes Warren Wiggins who is washing Waldo Woo.
I'd ask what counts as a "legacy system". We all have our "industry standard" definitions, a Burrows Mainframe, data stuck in an M database, etc. Is that what's really discussed here, or might it just be an old hand-coded file format, which can be moved to another kind of file server?
By moving everything to the cloud you're not eliminating problems, just making them someone elses problem, and enabling new ones to crop up.
Be careful of what you ask for, you might just get it.
Yikes. Proof that Ageism goes both ways.
'far too much influence'
You have an objective measure of this value?
Please share.
deleting the extra space after periods so i can stay relevant, yeah.
One year to move over 200 systems? Its an impressive target if he can make it. I know governments tend to move less like a Jackrabbit and more like a narcoticized slug when it comes to shedding legacy systems. I worked for a city government that had over 3000 legacy systems. They wanted fewer application and fewer servers. Their initial goal: one app. per year. A revised goal was one app. per month. There were change teams, change team documents, current app./new app. difference audits, needs identification groups, you name it. On the other hand I worked for a boss (he was team leader, I was the team), who spent about 3 months coming up with a mission statement for our team. Didn't do anything else, just 4 lines framed on a wall: 3 months. That's 1 1/3 lines per month. Have to give credit though: there were more than 4 words per line, so that's an entire word per week!
After it's a year overdue and 200% overbudget and everybody is completely blindsided by the fact that you can't quickly and trivially reimplement mature software systems, we'll hear a different take on this story. Of course, we get to pay for all this. *sigh*
Bray, who is one of the youngest CIOs across the federal government, shares his plan for updating those systems to a cloud-based, common data platform, that's "ideally open source."
Read: "ideally free in price".
Their entrenchment is the only thing with the power to bring down the empire.
Their entrenchment is the empire.
No government agency should have longtime employees. It's supposed to be a service, not a career.. It's these oldtimers that are making all the problems we endure. The bureaucrats are the "secret government".. We must purge them completely every few years.
So right about the time people start getting good at their jobs we should fire them all? If your goal is to ensure that the old stereotype of government being incompetent at everything gets reinforced, then that is a great idea. If you value things like, you know, competence, this sounds like a horrible waste of my tax dollars. And in many cases I trust the bureaucrat to do a better job than the politician who has honed everything to a 10 second sound bite without any real substance behind their ideas. At least the bureaucrats have skin in the game, in that they have to implement the policies.
namely, outsourcing all the equipment and control. make no mistake, OctopusCo doesn't suffer joy in the cubes, and doesn't care a damn about whether the work gets done. all they care about is the gaps in the contract. the way I'd look at this is, ramp up the cloud replacement, work in parallel for a while, and when it's proven, come in one night and pull the big switch on all the rusty old big iron.
if this is supposed to be a new economy, how come they still want my old fashioned money?
He'll be screwing up some other agency before the cost details get exposed. Maybe he'll have learned what the cloud actually is at some point, though. Doubtful. But maybe.
I can apply buzzwords and promote synergies by empowering individuals to maximize their unique contributions. My team even volunteered overtime during the holiday season, because they were so positive about our project. It wasn't because they were afraid they would be pushed out of their jobs by a CIO whose eager to ship everything he can out of house.
I guess he did okay at the CDC and hey, if it saves money, great, but who cares. Just do your job already. I'm sure the pay scale isn't that bad and the benefits are pretty awesome.
I would first review all of the system to see if they are still required. Then document data and methods. Then see if there are any low lying fruits that can be brought into the modern world. The goal should be to migrated all of it to a Web Interface and a unfified relational DB.
Is it necessary for the information stored in these systems to be online to be infected, spied upon, hacked. defaced. changed, deleted or held for ransom? Sure, replace the aging hardware (even if it "would take quite some time and would just reinforce stovepipes") but if the information on these systems are only internal documents, then keep it that way. I find the whole idea of "cloud storage" to be too dangerous to use unless data stored there is already "public" knowledge (like last weeks newspaper). Cloud storage is not your friend.
What's secret about the bureaucrats? They aren't a secret government they are the government. They handle micropolicy while the public debates macro policy. Do you want to have 20k public debates about water quality levels for each of the 20k lakes or one overall policy about water quality that then gets applied?
Wow. This will be ECSS all over again. You have nearly an impossible task. However, system integrators will be lining up to submit their RFPs.
We'll pull the plug after about 10 billion.
He "plans" to do it, he hasn't actually done anything. Boondoggles like this get started and never stop.
Surely, they can't all be mission critical? I bet most of them are upgrades that emulate the functions of the older systems, meaning there is a massive amount of duplication.
Historically speaking, government modernization tends to fail. There's the problem of duplicating complex and forgotten business processes. But that quickly becomes mismanagement: No-one wants to implement 2 IT departments and parallel change-over and extensive QA testing. Then there's scope creep, which is worse on a government contract and twice as bad on a modernization project. Next, governments are quick to prolong mismanagement with a 'failure is not an option' mind-set which also wastes time and money convincing bureaucrats the modernization process hasn't failed. Lastly, there are politicians demanding their electorate go in the pork barrel and bureaucrats having turf wars about control and job security.
Legacy systems have a few pros and cons, the ideal response is to evaluate the cost/benefit review, and availability for changes
Issues to evaluate
- Is this a specialized 3rd party product/hardware (may be restricted to vendor eg. ATM's were for a long time OS/2 well after IBM ceased producing OS/2).
- If it is 3rd party, do they still exist, or is there a similar product available.
- Is there specialized hardware requirements. (you may have no problem with a video card, but you might have a problem finding hardware for Wind tunnel Data collection)
- Has the amount of data processing increased/decreased.
- Is the service being utilized less and less.
- Do you have the manpower to handle the existing infrastructure for the proposed life expectancy of the product (if there are 3 people in the world that knows the system, and they all disappear, you may have a problem maintaining the system.
- Will a prolonged outage caused by system being unavailable due to the age, cause a serious impact.
- Is there a good justification that the changes will out weight the current value.
Pros
- business logic have been captured, and generated
- system has been optimized for the task
- known output
Cons
- Skilled labour, the languages, or hardware may not have limited and aging employee pools. (FORTRAN and COBOL are good examples)
- Increasing costs, technical people and/or parts become more difficult and costly to obtain. Some replacements may have to be custom made even
- Existing hardware could be slower
- Unsupportable protocols (eg SNA or Banyan Vines)
- Security, system may not have been patched for weeks, months or even years.
"Over time, this will allow us to turn off the 207 different legacy systems, and give us one common data platform that maybe has 207 different processes interoperating at the data layer on that platform. "
One process per system? Has this guy even worked in IT before?
Queue excuses along the lines of, "We vastly underestimated the size and complexity of the individual systems." in 3, 2, 1....
At least he has some app dev experience. Even it was developing a GUI....
https://www.linkedin.com/profi...
It's the software circle of life. The old team leaves and the new team comes in, looks at the software and says "Oh my god this is a giant turd! It must be replaced!" Well funny thing about that, usually a turd in the hand is worth two in the bush. Try not to think about that statement too hard. Anyway, my point being, the lucky ones get shot down by management immediately. The less lucky ones promise a shiny new future, end up mired in requirements and are quietly put down after a couple of years. The REALLY unlucky ones deliver completed software which doesn't offer as many features as the old software and has a ton of bugs the old software fixed over the course of two or three decades. By the time they get all those bugs fixed and features added, their software has morphed into a giant turd that must be replaced.
I'm trying to teach myself to set people on fire with my mind... Is it hot in here?
We must purge them completely every few years.
and replace them with corporate types like Wheeler! yes, regulatory capture at its finest.
mfwright@batnet.com
"Legacy" means it is old, but does it works? One of IT golden rule apply here: "If it's not broken, don't fix it"
Oh my god! A fond memory of the old days! Where's Tub Girl! Where's a Beowulf cluster of those!
putting the 'B' in LGBTQ+
A lot of the other comments seem to assume he is going to turn off the legacy systems. Long term that might happen but what he said was they would get all the data in a common place as the first step
That's much more sensible. It won't be easy mind you
The bureaucrats are the worst of the petty dictators. They can have you shot for stapling two pages together. We have to remind them they are there to serve. The best way is to eliminate the stagnation of careerism. It just becomes to familiar and corrupt with cronyism. Flush 'em out! In fact the entire government should have to be redrawn every 20 years or so. Sunset every law on the books, including the big one. Enough of the Victorian thinking we suffer now.
Moving the servers offsite doesn't magically make IPv6 work. Legacy systems are the single largest road block for IPv6.