Symantec study shows disappointment cloud computing is not magical
For a while, even hard-bitten IT types hoped cloud could end some headaches, not just cover them up
A new study from Symantec shows cloud computing has completed the rite of passage toward technological adulthood: The majority of potential customers say they’re disappointed in cloud computing, but have not dropped plans to adopt it.
The cycle of hype from overinflated expectations to the cold bath of reality is so predictable and, in many cases, so dramatic, Gartner has made a whole business out of diagramming the rise, fall and rise of a statistically average, generically successful technology and sticking technology buzzwords in various places along it.
Cloud computing, as Symantec’s survey of 5,300 companiesand Gartner’s infographic both show, have crossed over the Peak of Inflated Expectations and is heading down into the Trough of Disillusionment.
That’s actually a good thing for cloud computing, which caused a good portion of the potential customer base in IT to confuse the two realities any technology presents.
The happy, shiny one that shows up in airline magazines, shows how much easier it will make everything for business units and end users, on whose desks would appear a big, red button they have only to push to accomplish a hundred times more than they could by lifting bales or toting barges (figuratively) all day.
The reality for IT was always that someone had to go find a big, red button that could do users’ jobs for them without getting them laid off.
Figuring out what technology that button should control, how it would work, how it would connect all the other parts of the corporation together in ways that were only possible before by having an actual person do the work (let alone figure out what that person did all day, how, and how much of it actually needed to be done), has been the challenge of the Preventers of Information Services pretty much forever.
Cloud computing made everyone think it would suddenly be possible to buy a big, empty box they could open inside the data center to make all the complications of information technology go magically away.
Most of the hype was because some of the real benefit of cloud rinses IT with the same holy water as it does end users. But not all of it, and not without a lot of work.
Plugging in to Gmail or Salesforce.com or any other external application built, maintained, tuned, secured and guaranteed by someone else is easy.
Plugging in to complex, expensive data centers someone else builds and maintains should also be easy, unless you want it to do what you want it to do, rather than just sit there, blinking its lights, spinning its disks and waiting to be told what to do.
Choosing, installing, customizing, securing and connecting data to applications you would like to sit in a cloud maintained by Microsoft or Amazon or Rackspace is a lot less work than it would be in your own data center with software you have to write yourself.
It still takes a lot of work.
Guess what? Technology isn’t magic; it’s mechanics
If you want to rent a whole data center instead of just space for a few of your own apps and servers, it takes just as much work to design, configure and implement as it would if you were doing it in your own data center. It may be cheaper and easier in the long run because you’re not spending nearly as much capital to buy the hardware, hire the data-center crew to maintain it or even commit long-term to much of the cost.
It will certainly be a lot less work than building your own data center, and may work better, more efficiently and with better performance than you can manage.
It will not stay up 24 hours per day, 365 days per year, forever.
It will not run without losing a bit of data, exposing you to a shred of risk of data breaches coming from either inside the organization or outside, or requiring that your technology staff spend a lot of time tuning the networks, hardware, securing the data, writing custom interfaces, queries and functions for end users.
The cloud will not eliminate the headaches or work IT has always done to make boxes filled with semiconductors do things that seem important to people who don’t even know where the servers live.
If you design it correctly, if you implement it well, if you train your staff to manage the resources that go into your system, cloud computing can make IT far more powerful without necessarily having to build or buy a lot of extra power.
It can let you push computing power at apps that suddenly become memory hogs, see where the heavy users of data or applications really are and move both to locations physically closer to them – slashing lag time without changing a thing about how the apps or data work.
Disappointment caused by contradictory expectations
Symantec’s survey shows three quarters of organizations are talking about moving to cloud computing in some form, though only 20 percent have finished even their initial cloud projects.
Eighty eight percent of those surveyed said they expected cloud would make them more agile.
Only 47 percent found that to be true.
Disappointing, unless you realize cloud is a large-scale technology that only works as efficiently as it’s supposed to when it’s spread out across a large area.
It’s a matter of scale.
More than three-quarters of Symantec’s respondents said their staff doesn’t have the experience to complete a cloud project.
No kidding. And only 20 percent of you had finished one so far?
This is surprising?
Eighty seven percent of respondents said moving to the cloud will either improve their security or leave it the same, but it was also the point on which respondents felt most insecure.
New technology; new problems. No surprise that the security issues aren’t fixed yet, but the basic technologies underlying cloud computing are not brand new.
Wide-area networking, virtualized servers, networking, remote-access, shared-resource applications are all common and well understood with IT world.
How to do them securely within a specific cloud environment is not.
Cloud itself comes up short in some ways; it’s too young to have developed the kinds of safeguards required for many companies.
But it’s not a complete beginner. Private clouds are easier because you can control access, but that’s why most large organizations are using external clouds only for non-critical applications.
They know they have to work out the security issues before the use the new resources.
Disappointment due to inflated expectations of IT
An astonishing number of high-level IT people seem insecure about how far they’ve advanced in cloud computing because there’s so much hype about it, they just assume everyone else is farther ahead than they are, according to Sean Hackett, Research Director at The 451 Group.
Talking to senior-level IT people at large organizations it’s clear most companies are still wrestling with how to build their internal virtual-server and virtualized-application infrastructures most efficiently.
Once they finish that, then they plan to spend serious time working on cloud, he said.
Virtualizing servers changes the structure of all the applications and servers in IT, because none of them can count on the others being where they’re supposed to be. Virtualization means mobility, not dependence on the location of hardware.
All the apps, networks, access lists, security, directories and database managers IT has ever built depend on physical location. Virtualization wipes that out and forces IT to figure out how to most efficiently arrange both the virtual servers and the traffic they draw.
Cloud will make that worse, and better. Eventually.
Disappointment that cloud is not a miracle
Every new technology is hyped as a miracle, but cloud computing managed to fog the eyes of even cynical, experienced IT people because its whole purpose is to behave as if it is a miracle.
Resources appear magically wherever they’re needed, data travels to find the questions it needs to answer, servers pitch in with whatever work needs doing.
Done right, theoretically, that’s all possible.
However miraculous that seems, cloud is not a big red button on IT’s desk.
IT never gets to have the button on its desk.
IT gives other people the button.
IT is the button.
Cloud looks like a button, even to IT, but it will only work if everyone, including the one who gets to keep the button, builds all the machines that do the work the button is supposed to do.
That means automating a lot of what IT people do haphazardly, while putting out fires and answering endless questions from business units.
IT has no idea what-all it would have to automate to make that red button work.
That’s why IT people are disappointed in cloud.
First they have to figure out what they do, then figure out how to automate it, then build all the systems to handle the automation, then make sure they’re all secure, reliable and cost efficient.
By the time everything is in place to make it work, they’re all going to be too tired to push the button.
Read more of Kevin Fogarty's CoreIT blog and follow the latest IT news at ITworld. Follow Kevin on Twitter at @KevinFogarty. For the latest IT news, analysis and how-tos, follow ITworld on Twitter and Facebook.