By Sramana Mitra and guest author Shaloo Shalini
SM: Yes, I think that is what IBM is trying to achieve. I spoke with Pat Toole and Rick Telford at IBM earlier as part of the Thought Leaders In Cloud Computing series, and they both talked about how they are trying to build up the integrated stack in a pre-configured manner. IBM is planning to offer such functionality to people who are trying to do private cloud deployment. Their customers are the large vendors that are looking for the kind of configuration that you have just described. I think that is where they are going to end up. Among the small and mid-size companies, say outside of this Fortune 500 list, I’m not sure if anyone is doing that soon. On the other hand, this configuration that you are describing is probably more a Fortune 500 or global 2000 type of requirement, as opposed to what smaller companies need, isn’t it?
DH: I think there is a case to be made for both. The reasons why I say that is because someone who solves this would enable the customer, on an application by application basis, to kind of arbitrage between whether they want to host it on-premise, in a private cloud virtualized environment, or off-premise from some service provider. If that hybrid cloud exists, it becomes the grease for adoption of a private cloud than for public cloud. During the dot-com craze, a lot of people moved applications to what were then known as application service providers, right? There were several application service providers that had a great product and a great service offering, but they were overextended. They didn’t have the capital, they ran out of their supply oil, ran out of their working capital or whatever, and they went under or were bought by a different company. At that time, we saw a lot of people who were very happy customers of such companies and had good relationships with ASPs for some applications, be they Oracle, SAP, or even smaller ones, and they suddenly woke up in the morning to find that their provider had run out of business or was bought over by some bigger provider and all the rules had changed overnight. That is a very uncomfortable position to be in. We are talking about our corporate data, which is the lifeblood of the company. Being able to have to ease into a public cloud offering by using the same technology, same tools, same offering within your firewalls and knowing its expansion capability – that is going to make sense to a lot people. If indeed there are cost savings and operational leverage to be had from a public cloud offering, that would be the way to get to it.
SM: Do you are foresee a movement from private clouds to public clouds in the future?
DH: Only if the public cloud providers have a mechanism to integrate seamlessly with private cloud infrastructure. If that doesn’t happen, then cloud adoption, for the most part, will be limited to what we discussed already. You know, software as a service and maybe some platform as a service adoption for niche applications.
But this wholesale movement to infrastructure as a service . . . well, if I can quote another Gartner statistic, this one I don’t buy it at all when they say that by 2015 or so almost 25% of all companies will own zero IT assets. I struggle to see that happening today. But the only way it will happen is if there is a mechanism for easy, confident, and measurable slow migration from on-premise deployment to public cloud infrastructure. Wholesale adoption is simply too risky, too challenging, and potentially so painfully disruptive to an organization that when things don’t go right, the 10%, 20%, 0r 50% savings with cloud computing simply won’t be justifiable.
SM: This is very interesting. There are some parts of the cloud computing story that are moving very fast. Most likely the number one area is CRM. CRM has moved very fast, and I think about 25% of the market has already moved to an on-demand CRM offering. What you are saying is that on the infrastructure side, you don’t see such fast movement.
DH: You can’t run a company just on CRM applications. There are so many other applications. Again, I’m telling you something you already know, obviously. This becomes the crux of the problem. I think e-mail will start to move quickly to public cloud consumption at all sizes because it is a defined application, it is well understood, and there are lots of high-quality providers out there such Google, Microsoft, and others that can do it for you. I think that there are many CIOs, me included, who would love to hand off enterprise e-mail to somebody else, just because it is such a distraction and such a beast to deal with!
I think e-mail is good example of what would come next. There are probably others, the mid-market financials on the Microsoft product line. I think those are probably good candidates. Well, these are all software as a service applications, but they won’t obviate the need to own IT assets. At some point you need to answer this question: If you take all of these applications and send them out to a cloud-based mode, even if you can find a cloud provider for each one of your applications, have you really solved any problem, or have you just moved problems from one pocket to another along with cost? If you look closer, you will realize that you may have saved a little bit of money, but then with cloud-based applications you have 37 different vendors to manage!