Thought Leaders In Cloud Computing: James Dunlap, President of Cycle30 (Part 2)

By Sramana Mitra and guest author Shaloo Shalini

SM: You needed to go outside to find a data center which deals with IT support for your subscriber base of 450,000, is that correct?

JD: Well, even as we had a subscription count in the 300,000 range and up, that was really the point when we realized that we could not provide for that number as an IT organization. It could be a bottleneck and affect the quality of service, from a data center perspective, as we continued our rapid growth plans.

We decided that we needed to be able to leverage other environments where we could very rapidly turn up additional development and test facilities. We also needed to be able to turn them up and down as necessary based on project work and demand. We are constantly doing work within our “order to cash” environment, whether it is new trouble management systems, new order management systems, or activation or provisioning engines for new products. In the traditional environment, we would have had a staff of database administrators (DBAs), network administrators, and infrastructure managers to deal with such requirements. We did have such a setup in the past since we were a growing organization.

As part of dealing with our growth, we did make a conscientious decision to move that infrastructure permanently into a split environment where it is in a hosted data center. For our hosted data center, the provider of the service, SunGard, was responsible for the DBAs and any of the infrastructure. We use their cloud computing facility as a flex facility each time we require a new test environment, a new development environment, or other immediate requirements that need additional computing space and equipment.

SM: According to you, what is the delta between moving this environment to a data center and moving this environment to the cloud?

JD: Frankly, I think the delta is all about timeliness. What we found is when you move into your data center, you are operating primarily on your own hardware that you have purchased. If it is a managed data center, then you would have shipped the hardware to a provider and they would have done the configuration and brought up the infrastructure in their data center managed environment. What we found was that we could turn up environments extremely rapidly within the cloud computing scenario in a time frame that was dramatically quicker.

SM: Who owns the equipment? Was that the determining factor for your move to the cloud?

JD: I think it is really who owns both – the ordering, configuration of the equipment, and availability. Is there a cloud computing environment that is already up and functional? Are you simply leasing it or purchasing time on it as opposed to having fixed data center equipment that you own, or even a fixed set of equipment that they own that is there specifically for you?

SM: So, we are talking about existing infrastructure that you can flexibly turn on or off depending upon your need.

JD: That is correct.

SM: There is still a lot of confusion about what is seen as cloud computing. As you made the decision to pick SunGard for your data center, to what extent were you involved in determining what equipment SunGard selected? Were those factors in your selection of SunGard as a vendor?

JD: No, they were not. When we had our own equipment in their data center; those were our own specifications and engineering and design, architecture, and so forth. However, as they came to the cloud computing environment, that was not really a factor. We gave them our requirements in terms of how many subscribers we would be bringing into that environment, how many environments we needed, and the size of each of those environments. We did not set out specifically what that equipment needed to look like.

SM: Can you share the specifications of the environment?

JD: The specifications of the environment really were around the number of environments. The number of test, development, user acceptance test environments, and so on. It was around the size of subscribers we would be loading into that environment. We had disk space requirements associated with it. Network connectivity was also part of it. In this case, we were in Alaska so we needed appropriate network connectivity, for example, in the Sprint path where we could flex immediately to larger-sized bandwidth if required. We needed it to be a fully managed facility. By that, I mean I did not have spare DBAs or network administrators to loan out to them in order to manage any of it. It needed to be a turnkey environment where the administration of that environment was part of the service. I needed the ability to flex it up and down in the course of a year according to business demand.

SM: Is the business demand that you are going to base this on the number of subscribers?

JD: In our case, we can really get it down to simple terms – that is, the number of subscribers.

SM: For these development and test environments that your have been moving to the cloud, what is the software nature of these projects? Are these components of your “order to cash?”

JD: Yes, they are components of order to cash.

SM: Does GCI develop these components and then deploy them in the cloud? Alternatively, are these third-party software solutions?

JD: No, actually in case of order to cash components, most of them, around 99%, are the ones that we have purchased from other vendors, but we are integrating them together.

This segment is part 2 in the series : Thought Leaders In Cloud Computing: James Dunlap, President of Cycle30
1 2 3 4 5 6 7