Brocade's Klayko sees long path to cloud

12.05.2011
Brocade Communications remains a networking specialist in an industry where networks are becoming just part of a broader architecture for many vendors. A longtime market leader in Fibre Channel SANs (storage area networks), Brocade acquired Ethernet switch maker Foundry Networks in 2008 and now offers to tie together all elements of a data center or broader cloud infrastructure. CEO Michael Klayko has led Brocade since 2005. Last Tuesday, during an all-day press and analyst event at the company's headquarters in San Jose, California, he talked with IDG News Service about his company's own infrastructure and how it plans to keep up with bigger rivals.

IDG: There's a lot of confusion about cloud computing. What can you say to clarify what's happening there?

Klayko: You get 15 to 20 different definitions, depending on who you talk to. I would say I have a private cloud in my data center today. It's highly virtualized, it basically abstracts the application away from the operating system, we've got a storage area network. From an asset utilization standpoint, I have a cloud.

It's a business decision. Let me give an example: From a virtualization standpoint, when we talk about my business-facing applications, most of them are VMware. From an engineering standpoint, all my engineers have used Xen. In the old architecture, we couldn't mix these things together. With today's architecture, we can. I'll call it the cloud, because we have shared storage underneath it [and] we use the same server base. So we've built something that solves our business needs. And I think that's the real issue.

Everybody today that I talk to is struggling with proper asset utilization. So this big thing comes out called the cloud, and you don't have to worry about it: It's elastic, it allows you to put applications anywhere. It sounds great, and then the higher you go up in the C suite, executives all like to talk about it. They ask, why do you have to buy all these assets? Why don't you just buy a service level and get that from somebody else, and just utilize it like a utility? It sounds like utopia. We have 40 different applications that we buy from somebody else, and that run our company. We have our own infrastructure, also. We look at it from a business standpoint: Can I get it from somebody else, utilize their infrastructure and utilize their offering faster, more economically and more efficiently than I can do it myself? To me, it's math. I don't get emotional about it, and my IT guys, now I've got them not getting emotional about it. And I think most businesses, when you really get to the core of it, are like that.

Everybody knows how to build a private cloud now. All the tools are getting there, and it's all hinged around virtualization. [Then] there is the element of public clouds and the benefit of public clouds. The secret sauce that we're trying to get to is, how you merge those two. If you're a retailer, and four months of the year, your volume goes [up], you have to buy your infrastructure for the peak volume. What if you only had to buy it for the average volume you're in the rest of the year, and then just went outside during those four months? To me, that's a hell of a business application. Today, you can't really do that ... because nobody wants to allow you to have that infrastructure sharing out there until you have a long-term contract. It's not truly elastic, because they want you to use it and then stay there. , for example, this Cloud ID technology, will allow that elasticity.