Sramana Mitra: In terms of where it exists, are you saying that you also provide data center facilities to host a private cloud implementation of your solution?
Jon Freeman: That’s correct. We have many customers today who may own data centers but also have either co-lo facilities or use service providers to host some of their technology. Because of that, you may find yourself hosting or putting some of your services within an external service provider, an IBM or a Savvis or a Rackspace, for example, on behalf of an organization that’s also using those services.
SM: Right. I got that. Definitely, there is a third-party data center option that a lot of people are using for private cloud implementations. But the question I was asking is, do you also have a data center, a micro data center where you offer private cloud implementations for your customers?
JF: Yes, we do.
SM: Okay. You are offering whatever configuration works best for your customers.
JF: That’s correct.
SM: And your professional services background really comes in handy, because you can customize whatever works for the customer in that sense.
JF: That’s correct.
SM: What do you have to do, from your own infrastructure point of view, to be able to support these different configurations? How large or how much capacity do you have to maintain of your own? What does your IT need to support this business?
JF: Well, first of all, we needed to standardize on a specific, what we call, cloud fabric. That’s what we call our MEAD offering. Our MEAD offering is the layer that’s a standard virtualization cloud fabric that allows us to rapidly deploy services on top of MEAD, and still allows us to do all the monitoring, instrumentation, and capacity management regardless of which data center we’re deploying into. Our initial work was to be able to develop a fabric that allowed us to do just that. The first two years of our work was actually building that. Then we standardized on a set of utility type infrastructure components that would allow us to add consistency and remove any of the complexity around mixing and matching a bunch of hardware, software solutions. Our goal and our data center approach has been to standardize on a standard set of systems on a standard fabric, deal with each of the services on a one-to-one basis, and then develop, in conjunction with the services that we’re building, a set of configuration management tools that allow us to manage the configurations of the software systems without having to manage each one individually.
SM: I remember a conversation I had with, I think, it was Rick Telford at IBM Cloud Services. He’s the head of cloud services, and they were talking about creating these stacks for their private cloud offerings. They sell the entire stack to the customers who want to implement significant private cloud infrastructures. You are, I think, saying the same thing, but within the context of your cloud identity management solution where you have an infrastructure of your own. You have designed the entire stack, standardized that stack. And when you have to put up private cloud infrastructures for your customers, offering the same functionality, you take that same stack and implement it for your customers, is that accurate?
JF: That’s basically correct. What Rick is doing with CloudBurst is very similar to what we’ve done with MEAD. We develop a set of templates that can be implemented, that are fully configured with a full set of services. The entire platform, from the hardware all the way up to the security services, is managed and monitored as an endpoint device from our security operations center in St. Louis.
SM: Interesting. What I like about that is you’re productizing your professional services.
JF: That’s the dirty little secret of why we got into this business, quite honestly. It wasn’t to be able to offer a cloud solution. It was to be able to develop a consistent way to deliver our professional services without each implementation being a bespoke deployment. You’ve hit the nail on the head. That’s exactly what the additional motivator was for us. We were, as a professional services organization, always at the mercy of the customer, delays on hardware procurement, delays on providing adequate staff, network configuration problems. Our goal was to alleviate as many of the things that were outside of our control in doing deployments.
This is exactly how we started. We [said], how do we commoditize something that for the past 20 years has been very hard to commoditize? We were told it was cool to be in the cloud. But we did it because we wanted to add consistency, maintain margins, and increase margins by driving out all of the variables we thought were not in our control.