Few vendors have gained from rapid changes in technology and the economy, as much as open source poster boy, Red Hat. As enterprises looked at reducing their overall IT spend in a declining economy, open source become an attractive proposition—and Red Hat rode the wave. Today, even as the technology world fawns over cloud computing, Red Hat is looking at a bright future in the new cloud order.

In an exclusive interview with Srikanth RP and Harshal Kallyanpur, Red Hat CEO, Jim Whitehurst details why the cloud can be the mother of all lock-ins, why the idea of version 2.0 of any software will be dead, and why the new-world IT order will be led by a different set of leaders.

You have been quoted by some of the press in the global media as saying, “The cloud is the mother of all lock-ins.” Could you explain.
While the cloud has the potential to be extraordinarily powerful, it also has the potential to become the mother of all lock-ins. It’s like the line from the Eagles’ song ‘Hotel California’: “You can check out anytime you like, but you can never leave.”

The whole concept of a cloud is to provide the customer with choices that could help him ease his pain points.Some vendors are touting cloud-in-a-box.
However, we don’t believe in this concept.

A ‘cloud-in-a-box’ is an oxymoron. With cloud-in-a-box, you’re buying everything from one vendor, and thus you don’t have a choice. Hence, we do not have an offering called ‘Red Hat Cloud Architecture’ but instead we have a ‘Red Hat Cloud Foundation.’ We not only work around addressing the pain points of running an application within a cloud—but also help the enterprise to move it to another cloud.

What are the most compelling issues with respect to cloud computing today?
Most CIOs are concerned about the lock-in aspect of the cloud. If your data is stuck in a proprietary database and your applications in a proprietary application stack, your enterprise is stuck forever.

The issue about writing applications to a cloud such as Azure is that, if the cloud provider starts charging more for usage, where are you going to move the application? The key question to be asked today is, “How much are you willing to pay for a workload five years from now, if you are locked into that cloud?” In a few years, when customers have a choice, they are going to switch to offerings that do not lock them in. For example, in their sourcing rules, the UK government has laid down a component that one has to explicitly look at exit costs. Can anyone specify, for instance, what the exit cost of moving an application custom-built on Azure is?

Please share with us your company’s progress on the cloud front.
The cloud will significantly change the landscape and our position in this landscape. It has certainly accelerated our growth. It is still too early to determine the revenues coming out of our cloud offerings.
Today, a cloud could be a massive grid of virtualized servers and that is a decent part of our business. Over time, as an open source solutions company, we are taking the market share from traditional application vendors with many organizations choosing to use cloud-based services. Hence if you are using Gmail or Google Docs, you are using Linux. If you are using Salesforce.com, you are using Linux. These clouds run on open source hypervisors, middleware, management tools and applications.

Every time an organization adopts a cloud or SaaS model, they are using more and more of open source. Our products are naturally cloud-enabled as they are open source. Our Cloud Foundation initiative is similar to a set of cook books and reference architectures. It is a stepby-step 150 plus reference architecture that can help enterprises build a cloud effortlessly. In short, it has everything that an enterprise needs to physically build a cloud. You could use it to build a cloud on top of VMware’s ESX, use RHEL as the cloud platform and even run Windows environments on it. At each layer, you can use the hypervisor of your choice, OS of your choice, and applications of your choice—be it JBoss or WebSphere WebLogic.

With the change in positioning from only a Linux OS player to an infrastructure player, how has the competition changed, and how is the market ecosystem evolving for you?
For us, the three big players in the virtualization and cloud architecture space today are Microsoft, VMware, and Red Hat. While Microsoft has the Azure platform in the cloud, .Net for applications, SQL for the database, we believe there exists only a small section that will go all out with Microsoft for its cloud implementation.

There will be a larger section of enterprises that will use open source Java for application development. The two big companies that have exposed the layers—right from the virtualization level to the way you develop applications—are VMware and Red Hat. While we do not compete with VMware and Microsoft on a productto-product basis, we do have competing offerings in virtualization, messaging and application servers.

However, our vision is that the approach to cloud or virtualized environments needs to be layered, modular and based on open standards so that the customer can be given a choice. With VMware, their hypervisor and management tools are closely bundled with each other and hence, they follow a ‘buy this from us’ approach. While we can provide all the components to build a cloud, we do not want it to be a Red Hat cloud. Instead, we want to give the customer the choice to choose products of their preference.

The reason we are the largest player in the Linux space is that we have an ecosystem-friendly approach that supports a huge set of hardware and application vendors. We followed this approach in the data center space. While organizations such as IBM and Intel have derived more profits out of their Linux deployments than we have, we’re fine by it. We are taking the same approach in the cloud by trying to enable an ecosystem of the likes of IBM, HP, Dell, Cisco at the hardware level, or say the BMCs and CAs of the world, at the management level. By taking this approach, we are trying to standardize, modularize and commoditize the cloud.

I don’t see Red Hat ever competing with Microsoft by offering a suite of products that compete with their products; instead, we will be the infrastructure provider for a SaaS vendor that provides a solution that would compete with Microsoft.

Thinking more broadly about how the market develops, I will use this expression by Wayne Gretzky, an American hockey player: “Great players skate to where the puck’s going to be.”

We, at Red Hat, are already standing where the puck is going to be.

How do you see cloud computing impacting both the hardware and the software ecosystems?
Software will go through a radical economic shift.We’ve always been selling via a subscription-based pay-per-use model. Therefore, the move to cloud did not involve a radical change for us. For a typical license-based software company, it is a massive shift. The cloud will culturally change these companies.

First off, a typical software company provides certain features upfront, gets paid for them upfront irrespective of whether the customer implements these features, and gets the business value. With the cloud environment, they know they will get paid only after it is used and the business value is realized. It massively affects their economics. Also, if you look at general industry statistics, you’d find that 80 percent of users use only 20 percent of the software features. Hence, if you’re not pushing new features, how are you adding new functionality?

Typically, software companies release a new version and stop supporting the old one. And the customer has to pay more money for features that he may never use.

I’ve had conversations with hundreds of CIOs asking them, “Have you ever been forced to integrate a new version of a software with features that you may never use just because the old one would stop being supported?” I’ve got a 100 percent hit rate with each of them saying ‘Yes.’

With the SaaS model, the software vendor cannot really say, “We’re upgrading to a new version since there’s nothing to upgrade to. At the most, they are adding new functionalities to what is already available. It will be much harder for them to tell existing users that they will be charged more just because these features have been added.

If you look at our model—and it has been eight years—we have never raised our subscription prices. A traditional software company keeps milking its installed base of users by gradually increasing the maintenance revenue and through ‘upgrades.’ At some point, the customers won’t even look at the new innovation—they would be worried only about the increasing costs. In this new world order called ‘cloud,’ it would get increasingly difficult to milk the customer base with maintenance costs and upgrades—especially if it develops into a model wherein the customer can just move his applications between different infrastructures.

Another piece of this new world order is that the idea of version 2.0 of any software will be dead. Expectations are not being set by the likes of Google, Facebook or Twitter—but by their users. You would not hear of Google Docs 2.0—but they are adding new functionalities along the way. People are getting used to very rapid innovation cycles, and this idea will hold true for a custom application or an ERP application.

In the next five years, do you see more than 50 to 60 percent of software being delivered via the SaaS model?
It will be slower than that. People think Moore’s law is the driving force behind technology. By far, the most powerful force in order of magnitude is something called inertia. People are still running applications developed in the 70s in languages like Cobol, and the people who wrote these applications are dying. It’s hard for even us to tell a CIO who has been running an application for that long, to suddenly migrate it.

If you look at factors such as the number of people whose livelihood depends upon traditional delivery models, change management, and the inertia around it, it would take longer for SaaS to grow. It may grow at double digit rates when technology is growing at the rate of five percent. Will it double and triple every year? I don’t think so.

Change is hard; inertia is extraordinarily powerful. Hence, almost any industry is changed by a new set of incumbents and new-generation companies such as ours, Salesforce and Google, with new cultural models. And after all of us have become big established players, some other vendor will come in with a new way of delivering software and this cycle of change will continue.