Practical Analysis: The Slog Toward Private Clouds

Last week, I wrote about results from our cloud computing ROI survey; this week, we have the results from our private cloud poll. We asked about plans for the internal IT infrastructure and how it will mesh with external cloud resources, so opinions about software as a service are largely factored out of the responses to the latest survey. As it turns out, so is a lot of the maturity of thinking we saw in last week's survey. I don't mean that in a pejorative way: SaaS is simply better understood than the other forms of cloud computing, including internal private clouds, and that comes through in our survey.

Whereas 18% of survey respondents report using SaaS apps and another 31% say they're extremely likely to use SaaS, no other form of cloud computing shows even 10% using it or more than 20% extremely likely to use it in the near future. But more striking are the free-form comments we received. While a few commenters are quite certain about their attitudes on private clouds, most use hedge phrases like "might use," "could help," "too vague," "toying with," and "testing stage."

Those not heading down the private cloud route cite the usual reasons: No perceived business need and uncertainty about security top the list. In fact, it's the uncertainty factor that's stopping the vast majority of respondents in their tracks. Let's face it, it's an ongoing battle to get security, performance, reliability, auditing, and availability all right. Once you've got a handle on those things, you aren't going to give them up easily no matter the promise of the technology of the day.

Even when the benefits are undeniable, it's still a good idea to go slow, a lesson that should have been relearned with server virtualization. Boosting server utilization by a factor of 10 and consolidating lots of underused systems is a great thing, but the virtual server sprawl that quickly followed is a management nightmare. Then you spend a good fraction of those wonderful savings on tools to manage it all, only to find they aren't mature yet. Moving fully to cloud computing implies virtualizing storage and networking and putting a healthy bet on the capabilities of those not-so-mature management tools. What could possibly go wrong?

Here, however, the benefits are at least somewhat deniable, at least as perceived in our survey. The main business reasons for moving to private clouds are lowering ongoing costs, reducing capital investment, and accelerating delivery of services. At the same time, in order to get to private cloud computing, 48% say they'll need new servers, 45% new storage, 37% upgraded networking, and 35% upgraded security. Unless you're planning to lay off a lot of staff, something doesn't add up here. In that same question, less than 10% see the need for different identity management, application modeling, and user access systems, but you'll probably have to change those out, too, particularly if you envision any sort of hybrid cloud model.

Sure, saving money is important, especially since we tend to be running some very inefficient systems, but if that's the only goal, it's going to take some time for an entirely new infrastructure methodology to pay off. Keeping an eye on business agility and providing new applications is far more likely to pay off in every sense of the word.

Art Wittmann is director of InformationWeek Analytics, a portfolio of decision-support tools and analyst reports. Write to him at [email protected].

To find out more about Art Wittmann, please visit his page.

More than 100 major reports will be released this year. Sign up or upgrade your membership.