Blade Logic CEO Sees $10 Billion Systems Management Market

Looking to make a name in systems management, Blade Logic has been quietly growing at a steady pace for the last four years thanks to growing interest in lowering IT costs. But with the advent of blade centers and a new wave of distributed computing, Blade Logic CEO Dev Ittycheria says this space will soon mushroom into a $10 billion market. In an interview with CRN Editor In Chief Michael Vizard, Ittycheria talks about what economic and technical factors are driving a need for a new approach to systems management at customers such as HSBC, Walmart.com, Priceline, Time Warner and Putnam Investments.

CRN: One of the areas that creates the most cost and general inefficiency is server management. Why are things in this space so chaotic today?

ITTYCHERIA: If you think about the day to day lifecycle of a data center, there are typically five basic things that happen in the data center. One is there's a set of people that define what are the standards around the policies that they should have in place for best practices, and the standards in terms of how we build a management infrastructure. The reality is that whatever they define as standards are unenforceable because there's no way to automate those policies because they're very manual.

The second challenge is that the provisioning tools that people have are very inflexible because they're designed around images. Every time you change hardware, you have to have a new image. Every time you change an application or a configuration update, you have to have a new image. It's inflexible and frankly unusable for changes that occur very frequently. They're not really designed to manage complex changes. They're not designed to make very fine-grain changes. And they don't have the sophistication of being able to simulate a change before it occurs to see what may go wrong, or if the application work, or be able to roll back a change gracefully if something goes awry.

The fourth area is the whole audit and remediation area, what we call the compliance management area. Many people just do compliance management on things like patches and like higher-level things like files. But there is no integration to provisioning and change automation tools. If you do an audit and find out that one server is missing a set of files, or configuration files, there's no way to feed that back into your provisioning tool to make sure that the server gets provisioned with those missing files. It's very ad hoc. It's very manual. And it's very error prone. And then in terms of reporting, you only get basic hardware or software information. It's very incomplete in terms of provisioning change and compliance history.CRN: Where does Blade Logic factor into all this chaos?

ITTYCHERIA: People want to abstract the complexity of managing all these low-level configuration items by having this notion of service-oriented management. What that means is basically composing all your configuration items into a template or onto some sort of service model. So rather than provisioning 1,000 files and configuration files, you can provision this template that represents your Apache service. Then you can provision Apache. You can change Apache. You can update Apache. You can do compliance on your Apache template to discover misconfigured servers and you can remediate against Apache. It's basically designed to align the IT organization to the needs of the business, rather than just being this kind of chaotic world that people live in. Going forward from there is this whole notion of enabling utility computing, which is basically policy-based provisioning and remediation. That's the Holy Grail.

CRN: Over that last few years there has been a fair amount of talk about autonomic computing in the server management area, but very little of that technology seems to be taking hold. Why do you think this is the case?

ITTYCHERIA: The reality is that not many companies today are building data centers and using tools to do autonomic functions. The reason for that is that there's so much chaos right now so to go from this world of chaos to an autonomic environment is somewhat unrealistic. What people are trying to do is first just have much more visibility and control over what they have by starting to centralizing knowledge and define policies. Basically, they are imposing order and control. The next stage is moving to what we call the notion of service-oriented management, which is being able to manage and align your IT infrastructure and operations to the needs of the business. Then the third phase would be moving to more of a policy-based automation environment, such as a utility computing environment.

CRN: Could you build those tools today?

ITTYCHERIA: Customers don't want it. We have a prototype of a policy engine. But the reality is that if you don't have tires on a car, you know, you don't need a jet engine strapped onto the car. You need tires, you need a wheel, you need brakes and so forth before you go and use a car.CRN: What kind of momentum are you seeing this space?

ITTYCHERIA: The company was founded in 2001. In the last four years, we've grown the business on a compound annual basis of about 175 percent a year. We were basically breakeven last year and today we have about 100 customers across a variety of different vertical industries.

CRN: How big a market opportunity do you think this area represents?

ITTYCHERIA: We think that the market is conservatively at $10 billion. These are IDC numbers, not ours. To get that number, IDC says that there are about 20 million servers in production today, going to about 30 million in about two to three years. If you discount 20 million because they are small shops that will never need a sophisticated solution, then there are 10 million servers. Then you apply a price per server of roughly $1,000 per server. That's $10 billion. And then if you discount that by 50% again, just to be safe, that's $5 billion. And you're seeing it grow at roughly 60 to 70 percent compound annually, so you still see a $10 billion market in a couple of years. The market's huge.

CRN: When will this market take off?ITTYCHERIA: This whole notion of change and configuration is going to start getting a lot of buzz this year. Two years ago, the buzz was on utility computing. Then people realized that was still a lot of vapor and not much reality. Then last year, the talk was about patch management because of Microsoft's security problems. I think the pendulum has kind of found its happy medium now, where the problem is around configuration management for servers. This is probably one of the last places in technology where there's still a lot of manual process. I think you're going to see a lot of emphasis and focus on building a lot more automation and the management process of the data center.

CRN: Why is this whole area more important today?

ITTYCHERIA: Ten or 15 years ago, machines were viewed as expensive and people were viewed as cheap. You spent millions of dollars on computing gear and the labor, especially management infrastructure, was the rounding error. Today, it's actually the reverse. For anyone who has 100 servers or more, the inherent management costs of managing those servers is so high, with the labor that you need to service and support that infrastructure, it makes far more sense to use some automation tools to do that for you, rather than having to do it with just people. And if you do nothing, you're just going to have to add staff, which we just recognize is just cost prohibitive.

CRN: What is your go-to-market strategy?

ITTYCHERIA: Reach is our biggest problem. We use our partners where appropriate to expand our reach and footprint. Once you get the right person who has this problem, invariably we win a lot of business. Our average deal size is north of $300,000. We have about six or seven partner today and we're just kind of building out a channel organization. We need a distribution channel that can go out and create demand.

CRN: IBM and others talks a lot about this space. How do you stack up against them?ITTYCHERIA: IBM's been claiming to have provisioning for the last four or five years. Basically, they're trying to repackage Tivoli, which was a great solution for the client/server market, but completely inappropriate, and frankly very long in the tooth, for the distributed computing environment. The reality is that this is a very disruptive technology. We think that there's an opportunity to really take the market. That's reflected in the value of companies like Opsware, which is one of the most highly valued software companies in the public markets. What the market is saying, they're not necessarily valuing Opsware on financial fundamentals. What they're valuing it on is the fact it is a disruptive technology. Companies like HP, IBM, BMC and CA are at risk because if there's a discontinuity in the marketplace.

CRN: The general perception of these tools is that it takes 10 guys in lab coats to deploy them. Is that changing?

ITTYCHERIA: That's one of our differentiators. Opsware's approach is very services intensive and so is Tivoli and CA. The ratio of service dollars to licenses has always been very high, in the range of 5:1 to 7:1. Our services dollars are about 10 percent to 15 percent of our total revenue. One of the things that distinguishes us from our competitors is you can get very quick time to value. The way we've build the products is that as soon as you put an agent on the server, it automatically then scopes the configuration in real time. Other solutions require you to provision that server using their system, so that their system can discover what's in that server.

CRN: Can you integrate into existing console management systems?

ITTYCHERIA: We have a rich set of APIs and we obviously recognize that we're not selling to green-field environments. Being able to kind of be a good participant in the larger systems to management ecosystem is important. Typically the integration is not so much at the console level, it's more integration with workflow tools, like Remedy or monitoring tools. The monitoring tools do the find and we do the fix.CRN: Given the fact that you manage a lot more than blade servers, why did you call the company Blade Logic?

ITTYCHERIA: The reason why we call the company BladeLogic is that we think over time every device in the data center will become a blade. And the intelligence around how to provision, activate, update, change the services and the configurations of these blades will be based on software. The value is if your data center is going to migrate say, from two-way or four-way servers to tens of thousands of blades, are you going to manage your infrastructure the same way? No. The way you manage your infrastructure is going to very different than the way you managed it in the last five to 10 years. If your rate of change is going up, the number of devices is going up, the type of change complexity is going up, then you have to start building more automation into how you manage your infrastructure. That's in essence is the problem that we're trying to solve. And people are realizing they just can't use their homegrown scripts to manage the complexity and the rate of change they're seeing in their data centers today.