Wednesday, January 7, 2009

IDC's Al Gillen: Is a bad economy good for cloud computing?

Today's post is Part 2 of our recent interview with Al Gillen, program vice president of system software at IDC.

In the first part of our interview with Al that I posted yesterday, Al noted that virtualization is still on the rise, but more because of use cases like HA and DR than the traditional drive to simply consolidate servers. He saw systems management taking a much more important role, and virtualization serving as a "proof of concept" for cloud computing. The move to cloud computing is a transition that he says we will be continuing to talk about and work through "for the next 15 years." To me that’s a good indication of the impactful, fundamental change possible with the cloud. And the work that's still ahead of us.

In today's post, I asked him a bit more about cloud computing and one of the questions that's on everyone's minds: despite all the excitement on the topic (and, yes, the cloud computing PR train continues unabated in the new year, if my inbox is any indication), how will a dire economy affect all of this? His answer: don't expect people to suddenly make radical changes in dangerous times. They'll do what makes sense, after seeing some relevant proof points.

Here's the final excerpt from the interview:

Jay Fry, Data Center Dialog: Here's the "$700-billion question": I've heard two different schools of thought about how the current economic, um, tailspin will affect cloud computing. Do you see it speeding things along because of the chance to spend less (and get more flexibility), or slowing things down as IT becomes more cautious with any and all spending?

Al Gillen, IDC: This is a tough call. Cloud sounds good, but in reality, unless cloud computing really offers customers a seamless way to expand their resources, and does so at a lower cost, it will not be a short-term solution. Even if it does achieve these benefits in the first instantiation, cloud computing still has to overcome conservative concerns about things like data security, integrity, privacy, availability, and more. Corporate IT managers didn't get to their current job positions by taking excessive risks with company assets.

DCD: How long do you think the move toward cloud computing will take? What are the most important drivers and what's necessary for people to get comfortable with it?

Al Gillen: As I noted [in Part 1 of this interview], I believe that we will be talking about the transition to cloud computing for the next 15 years. Let's consider x86 virtualization, a market segment that VMware commercialized. VMware has been around for a decade now, and has had products in the market for about 8 years. Despite VMware's phenomenal success and billion-plus [dollar] revenue run rate [per year], the x86 server market remains only lightly penetrated by virtualization software. How much longer will it take for x86 server virtualization to become pervasive? Certainly yet another 5 years, at a minimum. It is unlikely that "cloud" can turn the industry upside down any faster.

DCD: You've heard talk from us and I'm sure others as well about internal cloud computing as a way to be able to help folks get the benefits of cloud computing without a lot of the current negatives -- and to do so using the heterogeneous IT resources they already have. What role do you see for internal cloud computing?

Al Gillen: Internal cloud computing represents a wonderful opportunity to get comfortable with the concept, and frankly, has the distinct potential to allow customers to lower their overall IT expense and raise their "greenness" using resources they have in place today.

On that note, I'd like to thank Al for being our first interview guinea pig. Despite the huge amount of interest around cloud computing and many vendor wishes to the contrary, what we hear from customers matches a lot of what Al says. The move toward running your data center differently comes in incremental steps, and those steps can't be rushed.

However, this sour economic climate is a great reason to champion IT projects and approaches that can save money and deliver results immediately. Who doesn't want to find a better, cheaper way to do things right now? (That's the reason we try to start out customer conversations with the payback/results requirements -- what is it you need to accomplish with your infrastructure management? Many people start with our IT ops and server power savings calculators to give them an initial picture.)

If you're interested in any of the research supporting Al's interview comments, you can get it (provided you're an IDC client) at their website. Frank Gens also produces IDC eXchange, an IDC blog with other, unrestricted commentary (including a lot about cloud computing and their predictions for IT in 2009) that Ken Oestreich pointed out to me. Al also noted that they have a Jan. 20 webcast to review those 2009 IT predictions that's open to the public. "We will talk about virtualization and cloud there, among other topics," he said.

No comments: