Slashdot Mirror
Who Makes the Decision To Go Cloud and Who Should?
Esther Schindler writes: It's a predictable argument in any IT shop: Should the techies — with their hands on their keyboards — be the people who decide which technology or deployment is right for the company? Or should CIOs and senior management — with their strategic perspective — be the ones to make the call? Ellis Luk got input from plenty of people about management vs. techies making cloud/on-premise decisions... with, of course, a lot of varying in opinion.
The first call comes from the technical people, and answers the question "Is the company technically able to move to the cloud, and if not what's required to get it to that point?". Once you've got that covered, then business can decide whether it makes sense to move and whether they want to invest what it'll take to make it happen. If it isn't technically possible it doesn't matter how much business wants it, and business can't make a determination about investing what's needed to make it possible if they don't know how much investment it'll take. You can't make a cost/benefit decision if you don't know the cost.
Honestly both should be involved. Anyone proposing a solution should be prepared to back that solution up when presenting it to others. I have seen Management make some really absolutely stupid decisions relating to software and platforms, but I have seen the Techie side do the same.
It all comes down to cost. Just like any other engineering or design issue.
senior management â" with their strategic perspective
This usually means that there is some financial input not available to engineering for decision making. Like kickbacks or Hookers & Blow.
Have gnu, will travel.
You mean you have stuff that isn't in the Cloud yet? What rock are you operating under?
LOL ... fuck that, let's be honest about it.
The perspective of a CIO isn't technical, or strategic, it's political.
He is carving out his own little niche and legacy, following the advice of the vendor he has most attached himself to, and in the direction which will give him the best story for the talking points he's been making.
He's unaware of the actual day to day stuff, doesn't give a fuck about why the shit he thinks is the wave of the future is completely fucking useless for the task at hand, and is really only interested padding out his fucking resume.
But don't for a goddamned motherfucking minute expect us to believe the CIO is making "strategic" decisions which in any way take into account the realities of the actual corporation, or pretends to understand the fucking current needs and requirements of the organization.
Like all C-level positions .. the CIO is an asshole, looking out for his own needs, looking to make himself look good, and doesn't give a flying fuck about the actual practical realities which exist.
Whatever particular kind of fucking hammer this idiot thinks is the kind of hammer is the hammer which will be applied to everything. Even if he's fucking wrong and clueless.
Strategic perspective my ass.
He may have a lot of industry experience, and he certainly has a perspective ... but don't fucking pretend it's a strategic one or an objective one.
It might be a tactical perspective for his own benefit, but the 'strategy' is in showing how awesome he can force a technology into the corporation, bonus points if he can make it seem to have saved money.
The long term costs and implications of this "strategic fucking vision" take years to understand. Bu which you've had several new assholes trying to impose their strategic vision.
The vast majority of CIOs are fucking idiots. And they don't "evaluate" technology. They just cram it down people's throats, and blame everybody else when their pet technology proves to have been the wrong fucking choice.
"Strategic perspective" this, asshole.
Those who truly run the business and call the shots are the ones who approve an organization-wide change of such great financial cost.
Like most things in life, decisions are made in collaboration. CEO wants "cloud", but when the CFO explains the regulatory risk, he finally gets it. IT folks might want to move "to the cloud" to pad their resume, but management doesn't want to spend the money on retooling/retraining/migrating (yes, there are costs beyond cloud costs in migrating).
People should talk, and see what makes sense. Like most decisions a corporation makes.
There is more than asking "tech vs title".
The better questions are:
1) Where is the data safer?
2) When it goes wrong (and it will) who will pay the bill?
3) What is the value to company of the data?
I work in world where major multi-year projects are under the way. If ANY of the information gets out billions of dollars will change hands. Product lines will be closed and others opened. Currently, even our email system is kept in-house. Corporate phones are blocked from using any voice assist, because the terms of service makes any information in those assistances not owned by my company. You come on property, your phone cameras have security stickers place over the lens. Removal of those stickers and you will be arrested for industrial spying and trade secret thief. NDA everyone.
So CxO on down, knows value of information and why those three question have the real meaning.
It's hard to explain who exactly should be in charge but IT staff should propose several solutions, their costs attached to it and the cost/benefits. There is never one solution and neither is 'cloud' a solution in and by itself. In the end, unless you're a small shop or require very small amounts of something, the 'cloud' is almost never the answer.
If you have 10 e-mail accounts, a hosted provider may look promising, but if you end up paying more for an e-mail solution in a month than you'd do buying your own hardware, you're doing it wrong.
Custom electronics and digital signage for your business: www.evcircuits.com
I've found that whoever has a position of authority in the organization makes the call, usually after having spent far too much time under the influence of a sales rep that lavishes them with meals or outings like to play golf.
Do not look into laser with remaining eye.
There's no hard, fast answer, although it would probably be popular around here to assume that the right place is with the Tech dept. This is certainly supportable; I've seen plenty of clueless administrators blinded by blinking lights and flashy fluff make architecturally very poor choices!
At work, we are a vertical stack cloud-based software vendor. We work with hundreds of clients and deliver a very excellent product that saves our clients $$$. Several times now, I've seen IT departments that have ballooned into inefficient "candy stores" for developers who are mostly intent on increasing their take of the organization's $$. It mostly happens because the managers at our client organizations aren't techies in any sense of the word, so they take whatever techno mumbo jumbo blurted out by the techies as gospel.
When the powers that be at the organization bring us in, and ask the tech department, they are almost universally ice cold to the idea of working with us, as their job is potentially on the line. Change = BAD! And so we see a fight while the corrupt IT department and the management duke it out. We've lost a few, we've won most. In any case, we often come in as little as 1/5 the cost of the bloated, internal IT department's offerings, while offering better service, better security, and strongly worded privacy and availability clauses.
So there isn't a right answer, you know? Some CxOs are clueless or corrupt. Some IT departments are similarly incompetent or corrupt. It all really comes down to "people are people".
I have no problem with your religion until you decide it's reason to deprive others of the truth.
whoever pays the bills MAKES the choice, usually, yes..
however, they are most often NOT the one best equipped to make that choice -- and therein lies the problem: bean counters and PHB fuck up IT decisions on a daily basis.
If you can't have a rational discussion between your architects (who are most likely just really senior guys slinging code) and your product managers (who are most likely just sales and account reps without a market vision) you are already screwed.
Shit. I just described my own company.
I'd vote CowboyNeal !
Knowledge is how to play a game, intelligence is how to win, wisdom is knowing what game to play.
First, it was outsourcing. Then came off-shoring.
IT managers have realized that people are not the commodity but the equipment is (although moving to the cloud also makes some people redundant). Having X number of computers brings almost 0 to a company's valuation. It is what's on they and how the data is used.
So it makes sense to move to the cloud! Well, sometimes until the dust settles and the bills start coming home.
What's interesting right now is the different types of clouds that exist. Amazon, Google, Microsoft offer more the building blocks of IT infrastructure, with essentially a price war going on. Oracle, HP and others focus on the higher margins with hosted applications.
Then, there's iCloud. A kind of personal cloud for hosting data.
It's pretty cool. There's something for everybody! In a way, today's technology and internet infrastructure enable computing as dreamt in 70s! (not exactly the same, like flying cars won't be exactly the same)
Technical management evaluates the options based on the requirements, and makes recommendations. BizMGNT shit-cans the recommendation in favor of whatever buzz-word technology is in vogue. Engineers implement the technologies chosen scratching their collective heads as to how this meets any of the initial requirements;. AND they have to engineer supporting infrastructure with no budget, and even less time to make pager-duty tolerable. At least that's how it's worked in my experience. Someone who doesnt have to answer the phone at 2am gets to decide.
CIOs/CTOs are hopefully technical in a company that has these needs, and they'll also hopefully consult with their people for these kinds of questions. If they're not capable or willing to do both, the company has more pressing concerns than where to host its stuff.
(Plenty of tech companies have bad CIOs or CTOs).
For every problem, there is at least one solution that is simple, neat, and wrong.
First, "techies" is like calling a manager a "flunky". Insulting.
Second, this has to be a 2 way conversation. The #1 benefit of "Cloud" is cost scope control - you can't customize it so business has a hard line instead of pushing IT for endless enhancements. As for "can or can't", provide managers rough estimates and high level todo's as to what it'll take to get there. There is never "we can't", there's "it'll cost this much" which will make then choke and chose a wiser path.
You want a serious 10+ year Ops verteran. And hopefully the company is established enough to leave the decision to the off of the CTO.
Finance. finance declares after 4-6 quarters of overspending on marketing and travel that belts need tightening and austerity is upon us. ragged edge carpeting and burnt out sections of office that havent seen a new coat of paint or replaced bulbs since the Clinton administration are passed over and the finger is pointed squarely at the company to stop renting beamers for golf outings. IT staff with crispy mice and rubbery keyboards are glowered upon and in turn the management steps in, begrudgingly, and does what management does to get the harsh glare of 'why do you have 5 monitors' off the team. Clouds are looked at, RFQ's are drafted, 30 minute powwows with the team are conducted and the reigning PHB compiles a short-shot list of the top 3 potentials to outsource the companies infrastructure to at a greatly inflated cost savings.
then its up to "the business" which in turn will glaze over as 3 choices are paraded out and the one with the most ad-buy in the seatpocket magazines on the flight to shanghai the CEO had to memorize for 17 hours gets chosen. After 3/4ths of the infrastructure is hauled kicking and screaming into the cloud, BIS screams bloody murder and keeps their mainframe while the exchange instance now shits the bed once every week. bitchcraft from the business about slow access, lost data, weird permissions, unplanned outages and lack of any discernable support chain are summarily compiled into a ticket system and ignored as IT staff shuffle together the last mighty years of their work into a functional CV and start shoveling tradeshow trinkets from their desk into boxes 'just to clean up a bit.'
6 months later half the team bails, the last guy who knew how to handle multifactor stuffs a notebook full of scribblings into a managers mailbox, and "the cloud" suspiciously gains 3-4 new employees with an impeccable history of providing excellent service and designing robust systems.
Good people go to bed earlier.
.
I've also worked for companies where such decisions were made by solely by the CIO. The CIO in one of the cases was not a technical person, and got the CIO position because of whom he knew.
The success of the project seemed to be directly related to the amount of discussion and information exchange among the "hands on" technicians and management (including business management).
The more discussion and information exchange, the better the outcome of the project.
Next relevant question: can the senior IT admin list the names of every single server the business has on premises off the top of his head?
Tell people that instead of saying "in the cloud" they say "on somebody else's computer" and see how that goes--
"We store the company's most important information on somebody else's computer"
"We control access to that data by storing it on somebody else's computer"
"We back up all our mission-critical information to somebody else's computer"
"Our data is secure because we store it on somebody else's computer"
Doesn't sound so good, eh?
makes the call.
Next silly question?
Sure, the person who pays the bills makes the call but they usually make that call based on cost. If it costs less in hardware/support/security/reliability/etc.. to move to the cloud, then it's usually a safe bet to do it. In most organizations, the person paying the bill would ask the IT department for costs and make the decision based on that. In the company I work for, we only have a few servers and by moving to the cloud we have a fixed bill, a reduced workload, no need to replace hardware, more reliability, etc... Once we found a good provider, there was really no downside. On a side note, our first attempt to move to amazon was a disaster and we have been much happier with stormondemand. The cloud isn't a single entity and each provider has different offerings and what works for one company might not work for the next so the whole vague "move to the cloud" is almost as useless as saying you need to buy a server or a network. You need to spec it out the same way you would spec out an in-house server, network device, or ISP connection and weigh all the pros/cons.
I'm the techie and I made the decision.
I found no compelling reason to switch to cloud services, but I do see the attraction of flexibly scaling up nodes to handle load as needed. It can be very cost-effective indeed if you've put in the work to manage the extra complexity, but hardware is cheap and developer time is expensive.
Right now we run a few servers with gobs of memory, and renting memory-heavy virtual servers isn't as cheap as buying the hardware.
It's not necessarily that so much as the head honchos usually have grandiose visions that are short sighted.
For that reason, and others, IMO it is best to follow the typical procedures one does in typical successful IT project management, which may sound bureaucratic, but it avoids disasters, thus these are pretty common procedures for a very good reason:
i.e:
1) Problem statement: Why is what we're currently doing insufficient?
2) Do a productivity gap analysis: Where are we now, where do we want to be?
3) Does the proposed solution provide for long term growth (so that we don't have to ask these questions again in only a few years)?
4, 5, 6, etc, etc,
But along the way, never leave out this important step: Consult with all of the stakeholders and make sure this solution works for them early in the design phase to figure out if this solution is even worth pursuing.
The stakeholders often include: Upper management, middle management, lower management, the techies, the rank and file employees, the customers, and sometimes the shareholders.
When you consult with them, what you're looking for are things like this: Does this new system work better than the old one? Does the new system make your job more difficult in any way? Does the system make your job easier in any way? What do you like about it? What don't you like about it? What else do you think you may need? (On that last question, make sure to have well defined scope limits early on to avoid feature creep.)
I remember at one place I worked, somebody higher up decided that they wanted to move all of the sales department to Microsoft Dynamics CRM some time before I first worked there. One day while I was asked to troubleshoot a problem with it, (and believe me, MS CRM has TONS of bugs) and I was totally stumped because this person's account didn't work even though his permissions were the same as everybody else's. After I did some investigation, I found out (and the management wasn't even aware) that every salesperson had this problem, only very few of them even tried to use MS CRM at all, so nobody actually reported it until this one guy happened to try it a few years after we had already supposedly been using it.
That's a classic example of when somebody implemented a new "IT system" but completely failed to consult with the rank and file employees to see if it's something they'd even want to use at all. It's also one of many classic examples of failed IT project management.
TL;DR: Basically, everybody who it applies to should have a say in it.
unless it's a fully encrypted solution where even the cloud provider can't see what the fuck you're doing.
True indeed... painfully true in many cases.
That said, remember that even though the CIO (and/or directors, etc) are easily swooned by vendors, consider this: One expensive fuck-up at the strategic level can destroy a career in less time than it takes for the CFO (or someone similar) to lodge a complaint in the boardroom, the first massive security incident, the first major outage... (and if you think the CIO is taking the heat for it, you're insane... that's why he drags a director or two into the process.)
That's why you don't (usually) see some CxO tromping up to a podium to announce "We're putting all our shit on the cloudz and it will be the awesomes!" without a metric shit-ton of consultation with (and agreement from) the rest of the C-level types. Oh *hell* no... first, he's going to drag the beancounters into it, and make them take part in the decision (if only to share the blame). Then he's gonna pile on the justifications (vendors will supply a lot of it, but if there is insufficient in-house justification, it's usually no-go), and make sure there's enough names on that thing to share the blame (but not enough to bury his getting the credit for it.)
All that said... no IT honcho worth his salt is going to do any of this without at least some input from a trusted sysadmin/developer/etc or two... if you think he did such a thing arbitrarily, then he's either flamingly incompetent, or that trusted person is definitely not you (or you're probably not high enough in the food chain, haven't sufficiently proven yourself, etc.)
PS: Fun bit about the whole deadline thing once a decision is made. He told the C-level guys 18 months as an ETA, but told the director to get it done in 6 months. By the time it filters down to you a day later, you got 6 weeks to complete it, and no weekends or vacations until it's done. :/
Quo usque tandem abutere, Nimbus, patientia nostra?
on the company and how IT savy they are. Some people still don't know what 'the cloud' means. In our organisation it's the CIO and management who make the decisions. However, it depends on the organisation. Ideally it's really a business decision but obviously technical experts should be consulted and no rash decisions should be made so that risks are avoided. The great thing about most cloud services is you can license a small amount of capacity to test and trial services before adopting. Sadly many people fail to do this.
I'm actually the guy making the decision for everyone from this point on. I've thought long and hard about it and have decided that nobody is going to use "the cloud" going forward.
Please forward my decision to your management no later than tomorrow, the cloud has been cancelled. I've also shelved the IoT, because it's a ridiculous and moronic idea.
Next week I'll be slashing a bunch of bullshit from the internet( ie. everything that isn't plain text), I'll let you know when to implement this change shortly.
Thanks for your cooperation.
In my view, nobody should comment on whether anything should be on-premises or not, if they are unable to distinguish between the word "premises" and "premise." Laziness of the tongue or of the quill is not an excuse.
It's senior management's call unless they request otherwise. Maybe the obvious on-prem location is closing abruptly, and senior management isn't allowed to announce that yet. Or maybe the cloud decision has already been made by a competent IT team of an unannounced acquiring company. Unsolicited technical objections might be, at best, a waste of time.
You can get a better performing, more resilient environment for less than your enterprise-class hardware. Plus you get free monitoring via Newrelic.
Save your business tens of thousands of dollars and get up to Aws.
How much is Amazon paying you?
Yep. Them dang magazines. Or them dang online "executive" websites. Often, "cloud" isn't an intelligent decision arrived at from collaboration between all concerned parties, as well-detailed in the earlier posts, it's "somebody's boss read a magazine article or saw something on the WSJ site," and next thing you know, the commandment comes down: WE MUST GO TO THE CLOUD, even though nobody involved in the issuance of the commandment even knows what the fuck "the cloud" actually IS. Or what it isn't.
Been through it many times. Every sortsighted headstrong exec who reads magazines hands down A Commandment:
We're Going Client-Server
We're Going Object-Oriented
We're Going Groupware
We're Going To Quality Circles
We're Going To Mobile
We're Going To The Cloud
They're a bunch of 5th graders worried they're missing The Next Cool Thing. They want the Pokemon, the Teddy Ruxpin, the Furby, THE COOL NEW TOY even if they get bored with it by the time they go back to school after Christmas. But in this case, everyone in the organization suffers for their fetish.
I mean, ask your friends.
"You guys putting things in the cloud?"
"Yeah."
"Why?"
"Management said we had to."
"Why?"
"Man, shut up. You already owe me a beer."
They know. You know. "Cloud" is just another fad. A new bandwagon. Some organizations can benefit from it. A lot cannot. But when you get edicts from on high, absent a rational review (again, as ably described in other posts in this thread), well, you're fuXX0red before you even start, because your organization is doing something because "somebody said so," and not necessarily because it's needed or useful.
---------------------------------------
Rotate the pod, please, HAL....
>better performing for less
Enterprise-class hardware? Maybe if you're overpaying for it.
If you owned the same hardware that amazon does, it would be cheaper and faster than running it on amazon's VMs. Faster for the obvious reason that you're running on bare metal and not inside a VM, and cheaper because Amazon wouldn't profit if their income from renting VMs was less than what they spent on hardware. Check the price of renting a big VM for 3 years versus buying the equivalent real hardware.
None of this is to suggest that there aren't other benefits to using VMs. Like you said, resilience and monitoring. Someone else's grunts are dealing with the hardware instead of you.
I'm just going to leave this here: https://www.google.com/#q=aws+outage+history
"A plan fiendishly clever in its intricacies"- Homer Simpson
Having worked my way up through every level, the biggest thing I've learned is "correct" is amassively subjective concept, based on value statements people at other levels don't see.
To take a deliberately simple case:
I would have declared a manager insane for buying Office365 licenses. After all, you can buy copies outright for less.
Except, as that manager, any savings I get are dwarfed by the pain in the ass of keeping licensing info. Some idiot loses the info and you're out far more than the difference when you have to re-buy. Or you don't re-buy and you're vulnerable to huge fines. Or you have someone dot every i and cross every t and you pay more for their salary than you save. Or Office365 keeps everyone licensed and demonstrably so.
Same goes for commenting.
Earlier in my career, commenting was slow. I could understand my code just fine without it. It was clearly readable after all. What idiot manager wants less productive code after I jumped through hoops?
Now I've paid the price of countless devs who write code no one else can follow. If watched countless more declare they have to rewrite everything because the previous guy who swore his code was readable wrote something the next guy swears is not. My perspective is completely different. I'd now rather each person codes a little slower so the company moves faster overall.
Who's right? Everyone has a good perspective but each is colored by the values that they weigh in.
I know my devs often think my calls are "wrong" because they assign different values to those I do... But I also know I've been put in the position exactly because I have the perspective I do. The best I can do is try to explain and help them understand, listening when they genuinely see something I've missed.
At a previous employer, I got to see this whole turn of events unfold [the wrong person deciding to move to "The Cloud"]. It went something like this:
a) CEO (non-techie btw) gets wind of "The Cloud"
b) SalesForce.com reseller somehow gets past the call screeners and directly to the CEO's phone.
c) CEO flies to San Francisco to a "DreamForce" convention to see Sting perform and hear Colin Powell speak and hear Virgin and Coca Cola sing praises to the platform.
d) CEO signs up for 3 years of SalesForce.com and a bunch of addons without consulting anyone
e) CEO flies back and tells everyone (and I quote: "OK everyone, I'm driving this car down the street with no headlights on, hang on, here we go!")
Needless to say I was out of that place not soon after. It was a real shame to see this "Cloud" technology forced down everyone's throats on a whim of the CEO, when he had absolutely _zero_ input from anyone else in the company (IT or otherwise). Especially when we had a really good system in place that just needed a few tweaks to make it perfect.
My friend who still works there now as to run around like crazy coding a bunch of APEX scripts just to hold things together. It's a sad, sad mess unfortunately.
The Kazakhstan problem goes like this..
You can't tell where your cloud computing is hosted. Any host can delegate to some other host.
So Vendor A with well paid shiny sales people sells cloud hosting.
Vendor A outsources the physical part of the cloud to Vendor B who is a bit cheaper with slightly less well paid sales people. Vendor A cashes the difference.
Vendor B outsources to vendor C in China who is cheaper still. Vendor B cashes the difference.
C -> D
D -> E
Repeat a few times until all the world's cloud computing it running on a server in a cupboard in Kazakhstan.
I should use this sig to advertise my book ISBN-13 : 978-1501515132.
I was going to say pretty much the same thing!
I've been a consultant for close to 20 years and I'm AMAZED by the decisions made by high level executives based solely on a sales pitch or based on the advice of a family member or friend that 'knows about computers and stuff'.
Do what thou wilt shall be the whole of the Law - Aleister Crowley
>better performing for less
Enterprise-class hardware? Maybe if you're overpaying for it.
If you owned the same hardware that amazon does, it would be cheaper and faster than running it on amazon's VMs. Faster for the obvious reason that you're running on bare metal and not inside a VM, and cheaper because Amazon wouldn't profit if their income from renting VMs was less than what they spent on hardware. Check the price of renting a big VM for 3 years versus buying the equivalent real hardware.
None of this is to suggest that there aren't other benefits to using VMs. Like you said, resilience and monitoring. Someone else's grunts are dealing with the hardware instead of you.
If you do it right, you should be able to guarantee some base load that will keep some computers busy. Buy those and run the system on it. Then use the cloudy servers to scale up as variable load happens. The cloud stuff isn't cheaper unless you are avoiding paying for idle hardware.
I should use this sig to advertise my book ISBN-13 : 978-1501515132.
The "techies" should submit a report, in writing, outlining the implications of a decision. No matter how much people hate writing reports, it does have a degree of accountability that casual consultations do not have. The writer is more inclined to provide both the benefits and the drawbacks of the decision as well as providing the rationale for approving or rejecting the decision. Documentation also forces accountability on senior management, since they have information upon which to base their decision. This is information that they have to take to their bosses if called upon.
This is not to say that the techies will agree with the outcome, but it can soften the blow. I have certainly written proposals for things that I did not approve of, but it was better than their alternative. (That original plans would have resulted in my resignation since they were planning to do something illegal. The alternative accepted their goals, but brought them in line with the law.))
It's interesting that somebody higher up decided to buy Microsoft Dynamics CRM, it has been my experience that CRM products, and especially Microsoft Dynamics, began as rouge deployments, purchased with knowledge or input from IT by some mid-level sales manager who wants to bypass IT and "cut through the red tape". Of course, only after several rounds of consultants have blown the budget and the disaster is complete is IT called in to rescue sales from the mess that they have made by trying to do an end run around the IT department. A competent IT department can usually restore some semblance of functionality to the average Dynamics installation, after devoting about 10 times as much work as should be necessary beating everything into shape. However, even in the best circumstances Microsoft Dynamics is still a hot mess that doesn't actually meet the needs of Sales who by this point have moved on to some "cloud based" solution and left IT holding the bag with the now abandonware Dynamics installation, at least until some other middle manager decides that since we've already purchased Dynamics we might as well try to wring some value from it and so another round of time, consultants and IT resources is thrown at the Cleavland Steamer that is Dynamics. In fact, I have yet to meet anyone who claims to actually be satisfied with Dynamics.
If you read the question to the end, you would know. The next question is then "and who should".
We used to have a Bill of Rights. Now, with the rights gone, all we have left is the bill.
There's a simple answer:
Get a techie into the position of CIO.
It's the C*Os job to make those strategic descisions and these should be based on experience. So have a techie as CIO and your MBA as CEO. And don't mix them up.
bickerdyke
Who should make the call to take any corporate data to the cloud? That answer is simple. The CSO.
Who usually does end up making the call to take corporate data to the cloud? That answer is unfortunately simple too.
It's usually anyone other than a CSO, with the end result being the justification as to why you need one.
I have done numerous migrations and helped out architect new services born in the public cloud. In 9 out 10 cases neither techs or CxO have a clue about how to use the public cloud.
Of course it is easy to show how blind management is, However it IT guys are not blame less.
IT has a history of the following bad behavior, that would make management want to find a way to slim its IT Staff.
1. Personal pet projects: This is often a business related project, however there are alternatives that may work better, however it IT worker is too emotionally interested in keeping it going, then giving it up for a better solution. Hanging on to the couple features that has that the others do not.
2. Attempts to make you "Irreplaceable": Sure that program your infrastructure you support is impressive, and perhaps no one else currently will want to touch it with a ten foot pole, and it is your baby, that is keeping the organization running. However in case of accidental death or injury the company is in a bad place, so they will want a better solution. And BTW just because people don't want to touch it, if they have to they can and will be able to maintain it, no matter how hard you make it.
3. Failing to project in the future: If they move to a cloud service, then your job is antiquated. However have you been future proofing yourself. Realizing the role you need to take after that particular feature moves away?
Now I am not trying to blame us IT guys for every stupid business decision... However you need to realize our personal bad behaviors do get noticed up, and influences business decisions.
If something is so important that you feel the need to post it on the internet... It probably isn't that important.
Often the Rouge deployments is due IT not being responsive enough or letting other departments behind while focusing more on others.
Those IT Guys they don't like us Sales folks, the engineering groups gets all the new technology while we get this old stuff. Well we have a budget we will subscribe to this web site that does what we wanted and tried to request for software for year, but was ignored by IT.
If something is so important that you feel the need to post it on the internet... It probably isn't that important.
You missed the risk assessment.
1. If the cloud company goes titsup/closes business, what is our fall back strategy.
2. What is our strategy should we find out NSA is spying on our customers and our provider is using the 'legal compliance' as their cover. (Applies to Microsoft, Amazon, Google.... all US companies, UK companies, likely all 5eyes and Russia and China)
3. What If the connection between our provider and our local server breaks. Surprisingly often, I live in a country that became a dictatorship and now port 22 SSH connections are blocked, Port 21 with STARTTLS is blocked on some providers. STARTTLS on email fails on some providers.
4. External hack. Can you ensure that the cloud provider complies with your local security laws? Can you ensure that *you* can audit *their* operations to the satifaction of *your* legal obligations/insurance/liability.
Just because vendors want to sell 'Cloud' stuff and market it heavily, doesn't mean it makes sense. Local servers are cheap, liabilities are massive.
Thats an easy one. This one happens like all the rest, as usual: Marketeers decide without asking the Techies. Techies have to solve issues in record time with no say.
When all comes crashing down, the techies save the day with the secret auto-backup they've been pulling off the cloud for the last 6 months.
We suffer more in our imagination than in reality. - Seneca
How a plan becomes policy
Ce n'est pas une signature automatique.
I always rather liked Bleu deployments... Rouge is so.... Moulin!
...gis sdrawkcab (usually not responding to ACs; don't bother posting as AC)
Some costs can be hidden from that deciding person. Or greatly understated.
I've seen that time and again.
...gis sdrawkcab (usually not responding to ACs; don't bother posting as AC)
. The cloud stuff isn't cheaper unless you are avoiding paying for idle hardware.
Er, that is the whole point of AWS. It is charged by load, so you scale up and down as required, therefore never paying for idle hardware. The AWS model wins because they extract 100% of every piece of hardware, while most on-prem shops would be lucky to get 10%. AWS is going to take over the world. You heard it here first.
I have spent hundreds (if not thousands) of hours cleaning up the carnage of #1
...if they have the sophistication to model the actual costs of the various options in a comprehensive way that includes deciphering the hazy costs associated with cloud hosting itself (availability options, CPU options, storage costs etc), premise costs (upgraded Internet access with true secondary path), migration costs (can some systems be just P2V'd to a cloud hosting provider or does it involve a platform switch and/or upgrade?), impact on staffing, general in-house implementation from a desktop perspective, etc, as well as longer-term contingency costs like deciding it doesn't work and needing to migrate back.
And then do the same for premise hosted systems, and then compare for reasonable lifetime of ownership (3-5 years).
Many IT people could probably do this themselves but Finance doing this would be smarter since they have access and understanding of other business cost dynamics. And it helps do it as in-depth as possible.
But it depends on the definition of finance and how sophisticated they are. My experience lately has been a surprising number of cloud adopters going back to on premise for a lot of things because they found cloud to be expensive and when they looked at what they actually ended up paying versus what it had cost them to do it on site found on site to be cost competitive and have flexibility that cloud lacked.
This leads me to believe that cloud pricing is very opaque and these organizations didn't do a lot of cost modeling, somebody heard cloud and some kind of $19.95/month sales pitch and mentally compared that to the last IT infrastructure invoice they paid.
AWS uptime of 99.9974%, down for a total of 2.41 hours over 20 outages across the globe in 2014. (I'm guessing most of their zones had 0 downtime) This was listing ALL downtime worldwide.
We use AWS, it was crap many years ago, I've had 1 server unreachable for about 45 minutes over the last 2 years. I wanted to check you link and found that is about average for them. Our downtime was told to us 40 days before it happened, we chose to do nothing and deal with the downtime.
So if you were implying its always down, your link doesn't back up your implication. Maybe you hoped no one would click the link and check?
I still think its a bad idea, but falsely quoting downtime makes me not want to believe you on anything with this topic.
Cloud is like outsourcing, as in a) it'll be a management decision that b) won't save that much money and c) will result in less efficient business processes. 'strategic perspective', do you mean what the CIO just read in the non-technical press.
There is no such things as "IT decisions". There are only management decisions. All IT gets are orders.
Corporate level risk decisions like this the final decision has to rest with a C-Level executive. If he is a good manager of course he will get input from other people first then make the decision. The eventuality that the corporation cloud service is hacked or compromised ... they will have to answer to the owners or the shareholders anyways.
Personally, I would be extremely wary of allowing any corporate data to be "housed" in a cloud unless they have deep pockets are can be held liable for damages caused by a major breach.
All I will say to this is; What happens in the Cloud, does not stay in the Cloud. Then tell the boss to ask Ashley Madison. I'm betting their site was in the cloud. They say something like "there's 38,920,000 anonymous reasons".
Everything is the cloud, they just renamed the internet to "the cloud" so that we could party like its 1999 and have ASPs (now SaaS) again. Virtualization is not new (the mainframe did it 30 years ago). Did you ever wonder why there are so many liberals in the "fashionable" IT companies (e.g. Google, Facebook, Apple)? Its because liberals believe that they know better for you than you do. Same reason why they are in the media - to have control. So, the idea of seizing central control of all computing is a fucking liberal wet dream. They've wanted this since the 60s and now that those hippie bastards are in control, they are forging their will into products and services to realize their totalitarian dreams. You think those MBA asshats from Harvard, Stanford, Yale are out to do anything but to bless us with their benevolent wisdom? Hell no! We're just a bunch of fucking sheep that need to be controlled.
You really want to know where this shit comes from? Subscribe to the McKinsey email lists for technology and the Harvard Business Review. Those places are the MBA meccas. The source of much of this nonsense.
In other words, good management should be smart enough to take everything into account. Without that, the company is rather screwed.
If management doesn't feel they can trust the people they hired to manage the IT side of their business, then either they have no idea what they are doing or they hired the wrong people.
Just like the people you hired to handle security, or accounting, or any other job in the company, it is the responsibility of management to make good decisions on who to hire at the top levels of IT. If they don't trust trust the opinions of their hired experts then they should get new ones.
So no matter what, management ignoring the opinion of their IT department on this indicates a management problem.
In the U.S., it just takes a well-greased palm to decide the best I.T. policy. The Panamanians have nothing on U.S. I.T. corruption in big companies and government.
The person(s) who have the authority to sign the deal, more precisely, are responsible for the decision.
Said person(s) are absolutely stupid if they do not lend very strong weight to the opinion of their IT department, however.
There's a behind the scenes push where I work to move all of the active directory part of our systems to the cloud under Microsofts control. We already have a hybrid exchange environment via office 365 with a large portion of accounts maintained in the cloud by microsoft.
Some services make sense to offshore into the cloud. You free up money by not having people dedicated to supporting those applications, and hope that outweighs any downtime.
But moving AD to the cloud? People complain when we add something to a GPO that adds a few seconds to the logon time, and we are expected to offshore our AD infrastructure to the cloud? That's putting a lot of authentication eggs into one basket relying on the internet. Sure we can cache credentials locally, but what happens if it's out before YOU cached?
Beyond that, knowing what I know about vmware and other virtualization technology, you are only as secure as the weakest VM on the host. Do you really want to put your important stuff at risk? There were at least 2 bugs I had to patch on RHEL boxes over the past few years, specifically dealing with Xen, but most VM products also had this hole. A user on joe blows hosted VM is able to break out of the VM and gain access to the other hosted VMs.
My nightmare scenario is a major break in the infrastructure removes one or all of the sites from the "internet". We can handle that fine right now and continue to deliver services. If we lost access to AD because it's in the cloud, we're only able to operate as long as we're caching credentials.
Cloud, proactive, synergy.... when you hear these terms from a consultant, quietly close your notebook and continue nodding and smiling until they leave, never invite them back.
You can't necessarily trust the techies either since these cloud projects offer the CIO the opportunity to cut "human costs" by way of layoffs. The techie's opinions are tainted by fear and/or the prospect of seeing their empire shrink. And your CIO knows this. You will find out how much he truly trusts you.
And yes I know that many of these projects do not lead to such productivity improvements but the risk is real.
So if you were implying its always down, your link doesn't back up your implication.
Yep, you got it. I was implying that AWS is always down. Always. No exceptions. Never running.
Good job.
"A plan fiendishly clever in its intricacies"- Homer Simpson
We used to call them System Analysts. They were seen as dinosaurs and nobody needed them anymore. They became extinct.
Since I do sell this, I'll give you what we say to our potential customer:
Is your IT currently or going to be following an OpEx model (meaning, do you want to move this to a monthly cost vs a huge captial expenditure)
If yes, then you have a possibility for an external Cloud offering, because there is no on-prem infrastructure, just ongoing (forever) monthly costs.
If that's not the model, don't choose a cloud offering. Some customers are looking to drive it this way, some are not.
Two, have you ever done anything else in the cloud before ? If yes, ok, maybe this is still an option for you. If not, then we advise you step back, and choose a small segment that you want to try this with, such as developers, before you attempt any earth shaking move like an entire company.
Three, where are your datacenters now in relation to your end users? If the distance is too far, then stay with what you have. Not every part of the world will a cloud be the right fit.
So, to me, the very first question is about OpEx vs CapEx, because if the boys (or girls) paying for it are not looking for a new model, this will never be successful.
Incidentally, by definition this is never a problem on the business side..
HANDS OFF!!, my pretty...
Cheap storage VM.
Of course it is easy to show how blind management is, However it IT guys are not blame less.
IT has a history of the following bad behavior, that would make management want to find a way to slim its IT Staff.
1. Personal pet projects: This is often a business related project, however there are alternatives that may work better, however it IT worker is too emotionally interested in keeping it going, then giving it up for a better solution. Hanging on to the couple features that has that the others do not.
Strangely, my experience has been that our devs test-driving "the cloud" has made them appreciate the level of support they get from the in-house IT even more. Almost without fail, every dev team has gone through a phase of "ooh! we can pop up a server without waiting on IT!" Followed by "This isn't *exactly* what we need, and we can't change it from the canned offerings," and "hey, IT support us! Sorry, we can't do anything for you past general advice, you have to work with $CLOUD_VENDOR." And eventually, they come back in-house where we can give them what they need, and have a team of subject-matter experts that can immediately jump on any problems.
Of course, we're fortunate not to have the IT dept staffed entirely by imbeciles, which probably makes a big difference.
...and by moving to the cloud we have a fixed bill...
...and by moving to the cloud we think we have a fixed bill...
FTFY
Cheap storage VM.
One thing that gets me in discussions across organizations is how poorly the "cloud" is defined. IT often has a slightly different definition of the cloud than senior management than end users than tech support (and so on). Are we moving email to the cloud, setting up a collection of virtual servers to run our custom apps on, using Salesforce, creating a hybrid solution for redundancy? Even in those situations, the motivations and concerns are often different.
Then there's the accounting aspect. Is the shift simply to move IT from CapEx to OpEx? Does the IT staff understand the difference? Has management worked out a 3 year forecast to make sure the financials actually work out?
When making these decisions, all major stakeholders need to be involved and represented. You need to look at it from different perspectives and make sure everyone understands those perspectives. Only then can you really make an informed decision. Yes, that's much more difficult than simply believing the sales guy, but for something as important as IT infrastructure, it's what you should do.
I work on the laboratory informatics/gene sequencing side of the world and these conversations are becoming more common. To help give scientists some perspective, I've putting together some blog posts that introduce all the different angles:
https://www.lab7.io/is-your-he...
Yes, it's a bit of shameless self-promotion, but it's also relevant to the discussion (and I don't want to just cut-and-paste it here :) ).
-Chris
Our organization recently decided to go with Microsoft Dynamics Online. We (IT) didn't really get much of a say in it.
However, what we did was use the initiative as a platform for our own move to expand the cloud deployments from just CRM users to all users.
We are now well on our way to migrating off of on-prem Exchange and a mixed Office 2003 - 2010 environment to an Office 365 solution.
So, in effect, we took an project spearheaded by sales and turned it into a benefit for all users.
My eyes reflect the stars and a smile lights up my face.
FTFY
How do you figure? We now have a monthly fee that is the same each month versus previously we had an initial outlay of several thousand dollars per server, ongoing costs of hardware repairs, and then another outlay a few years later for new servers. Yes, if you don't have a contract then the monthly fee can possibly go up but it usually just tracks inflation and is much more fixed and steady than owning your own servers. You can argue that it's more expensive, it's less secure, or a host of other things but I honestly can't see how you can argue that the bill isn't fixed. I can show you our bill for the past 2 plus years and it is definitely the same every month.
Somewhere in their billing structure is a gotcha. You haven't hit it yet.
It could be higher loads, it could be when you need something quickly.
Cheap storage VM.
Somewhere in their billing structure is a gotcha. You haven't hit it yet.
It could be higher loads, it could be when you need something quickly.
Why so pessimistic? Believe it or not, there are plenty of companies out there with straightforward upfront pricing that don't nickle and dime their customers and instead charge a fair price for their services.
Yes, if you need technical support for something that's not under contract, a company might charge you a reasonable hourly rate but believe it or not most companies aren't out to screw their customers. Especially where there is a lot of competition, it's hard to stay in business if you're always trying to squeeze your customers. It's a lot easier to make money if you just charge a fair price for a good service.
So what you're saying is, you're not a very good, or effective consultant? You get hired to give advice, and are regularly ignored in favor of J. Random Salesguy?
This isn't an either or scenario.
The tech team should be consulted to make sure any new solution is sufficiently compatible with existing ones. HOWEVER, they are just one of the stakeholders in the process: users will flat out reject anything thrown at them without consultation, business leaders may know about competing projects or business goals, and any major initiative is likely to trip over a handful of shadow IT projects.
The CIO needs to be the ultimate decision maker because they have the perspective to see all these things at once. They also have the authority to get rogue stakeholders in line, if need be. Without consulting the tech team first, any project is doomed to fail (or at least a very slow and painful success). Of course, that is in an IDEAL scenario. In reality, CIOs tend to be cowboys, other C-level execs tend to compete for authority and funds more efficiently, and cost-center mentalities can write off any IT project as a costly failure before it even starts. That's just getting started.
If you're having problems with the decisions your CIO makes, here's a trick I use - instead of proposing 1 or 2 solutions, propose as many livable solutions as layers of management will touch the proposal + 1. Everyone in the process likes to be a "decision maker" and will trim one option and send it upwards. By the time it reaches the CIO, it will have 2 options on the plate - which is exactly the type of binary decision that most C-levels like. This will get your proposal through upper management quicker, and still leave you with a decision you can live with. It works in many cases, but you will need the support of the other stakeholders, and of course YMMV.
In regards to the cloud - everything is a cloud. It's a matter of where the services need to live, and what amount of liability are you willing to coexist with. (Pro Tip: Liability can't be outsourced.)
Whoever has the guts to stand up to the marketing drones and tell them, "No".
"Nine times out of ten, starting a fire is not the best way to solve the problem." - my wife
I always rather liked Bleu deployments... Rouge is so.... Moulin!
When two different posters used the term "rouge deployment" I assumed it was some new buzzphrase rather than a typo.
To have a right to do a thing is not at all the same as to be right in doing it
As it should be.
I was going to say pretty much the same thing! I've been a consultant for close to 20 years and I'm AMAZED by the decisions made by high level executives based solely on a sales pitch or based on the advice of a family member or friend that 'knows about computers and stuff'.
Unless you're talking about some one-man band operation, "high level executives" do not get to make major unilateral investment decisions based on random hunches.
To have a right to do a thing is not at all the same as to be right in doing it
Did not Heinlein write about techies vs. management in The Roads Must Roll?
Tracy Johnson
Old fashioned text games hosted below:
http://empire.openmpe.com/
BT
While techies may know the technical aspects ALL THINGS in a corporation are management responsiblity by definition, economics and legal structure. To them, IT/CTO are merely advisors and implementers. That's it. If it makes sense to use the Cloud, then it must make sense for the business first. It's not merely a toy for IT or an empire for the CTO to create.
The Cloud is a majorly over-sold bandwagon right not. Just as much as people who imagine that "all IT should run through port 80" - there are such people and they are dangerous morons. Recently I saw that someone was actually contemplating an MRP system "on the cloud". It's when things reach this level of idiocy, IT people loose all legitimacy and trust!
I've lived in both worlds, and just as there are Pointy-Hair bosses, so too the Dilberts are Epic Fail also.