Sun's 'Open'-Door Policy
Sun's 'Open'-Door Policy
Sun Microsystems President and CEO Jonathan Schwartz and Executive Vice President for Software Rich Green run what is now the largest open-source-based IT infrastructure company in the world.
Schwartz became CEO of the 26-year-old company in May 2006 when longtime President and CEO Scott McNealy moved to take over as chairman of the board. One of Schwartz's first moves was to rehire Green, who had run Sun's Java and Solaris businesses-among other things-for 14 years, from 1990 to 2004. Green has overseen the company's final two years of transition to a company that develops open-source software, uses it as an entr??Â«e into enterprises, and then sells hardware and services into those enterprises.
Schwartz and Green sat down with eWEEK Editorial Director Eric Lundquist and Senior Writer Chris Preimesberger on April 8 for a wide-ranging conversation at the company's Menlo Park, Calif., campus.
Sun has made the transition to a legitimate open-source business model. What is the next step?
Schwartz: Open source is a means to an end. It's a mechanism to grow the broadest market, build the largest ecosystem, reach the maximum set of opportunities ... but just because you've reached the market doesn't mean you've effectively built the business. It's the difference between being popular and being successful.
In your estimation, is the open-source message getting out to the institutions, by and large?
Schwartz: They may not be the target demographic. Because, for 95 percent of the CIOs I interact with, when you say, "open source," they cringe. And they worry. And what they want to know is: Is there robust enterprise support, a high-integrity road map, great innovation and full legal indemnification? So that's what we talk about when we talk to CIOs.
And again, none of them is likely to join the community and submit a bug fix. That's the antithesis of what they want to do. On the other hand, if you talk to a developer, or a startup, or a developing nation, they care an awful lot about the fact that we have the highest-integrity open-source license, highest-integrity open-source community-the only real productive commercial engagement with open source in the marketplace.
Lots of other companies have assembled intellectual property and then put their brand on the top and said, "Look what we did." We actually put 11,000 engineers to work at Sun to go fuel those communities, seed those communities and drive those communities. The net result is we produce innovation-with brands like MySQL, OpenSolaris and Java-and then we produce products and technologies that allow businesses to run more efficiently. At the end of the day, that's what CIOs care about. They just want to run their businesses better, and they don't want to get involved in political philosophy or theorizing.
Effects of the Slowing Economy
How is a slowing economy affecting the company?
Schwartz: A slowing economy is great for free software. It's fabulous. Look, if your No. 1 driver of cost in your data center is proprietary software, the fact that you pick up MySQL simply by picking it up and putting the MySQL coffee mug on your desk, you could save $15 million or $25 million. That's what we're seeing across the world.
I was with the CIO of a very traditional financial institution recently. At the end of our meeting, I said, "By the way, we've just announced the closing of our acquisition of MySQL." The CIO looked at me, and she said, "Well, that's nice, but we really don't use MySQL here. We're a proprietary software shop." A very eager Sun sales rep was with me who had checked in with his buddy at MySQL and found out that this organization had downloaded MySQL 1,300 times in the last six months.
[The CIO] was stunned by that. A couple of their technology folks who were also there said, "Actually, it's the No. 1 database all of us use. It's just that we don't have a commercial license because we've been told we're a proprietary vendor shop."
So now we're in the midst of negotiating a license, and they'll wind up saving, like everybody else, $5 million or $10 million. And that, in a slowing economy, is a very helpful thing.
Are the old-school, institutional enterprise companies generally beginning to see the value of open-source software?
Green: It varies, but it always has the same ending. It's very funny. Last year we had a Sun Tech Day in St. Petersburg [Russia]. We had the event in a hockey stadium, the biggest place we could find. So I asked the audience, "How many of you are using MySQL open source?" About three-quarters or 90 percent of the people's hands went up. Remarkable! This is pretty far away, and it was a classic case of the overall adoption of open-source technology.
Wherever you go, it's [open source] in wide use. Like the years-ago growth of Linux or Windows, this is a bottoms-up trend that is unstoppable.
This [open source] is about unfettered access to technologies, the contributions of communities, and development by the masses that ensures the most capable, the most useful and the most secure platforms because it undergoes the scrutiny of the largest collection of eyeballs in the world.
What is Sun's strategy for selling the development of cloud computing?
Green: Sun is in a unique position as a neutral party to provide a great cloud computing platform for developers. Our view is that we have a series of open-source technologies, in NetBeans and others, that have already been extended to provide cloud computing interfaces. There are already extensions in NetBeans that allow you to direct your development efforts to deploy on the network dot-com cloud.
The other thing you can quickly expect to see from us, given the acquisition, is hosted MySQL services, which is an obvious and incredibly valuable platform, as we now have MySQL employees working under our roof. Who better but Sun, with that brand affiliation and expertise, to provide that as a database cloud platform? It's a high priority and is already under way. You'll be hearing a lot about that soon.
There's very little support for cloud computing right now. But imagine the possibilities of millions of NetBeans users, sitting in front of a development environment on a laptop that just has a deploy button. It conjures up all the working components, delivers it to your working space, and deals with either free or micropayment functionality, deals with the security interfaces, and does it out of the deploy button. No one but Sun can do that right now. That's where we're headed.
Schwartz: The cloud that I probably have the most respect for right now is the Amazon cloud. I think they've done a wonderful job of actually understanding what cloud services are, which is, you build a running operating platform, you give it to them, they provision infrastructure for you, and then you can run it. And you're not tied to Amazon, so the day somebody else comes in and says, "I have a cheaper provisioning solution for you" or "I have a better reliability meter for you," you can feel free to move.
What Brought Green Back to Sun
Rich, you were with Sun for 14 years, left to help startup Cassatt in 2004, then returned two years later in May 2006. Why did you leave, and why did you come back?
Green: Completely different reasons [for both moves]. There was a bit of controversy here. I was running Solaris and Java and was involved in the Microsoft [antitrust] litigation. It got to the point where I could see the end of the litigation. My last day at Sun was when we settled for the $2 billion. It was sort of, 'You ran Solaris, you ran Java, and you secured $2 billion for SMI [Sun Microsystems Inc.]. ... I'm going to Disneyland.' It was a time to do something else.
It was 2004, and it was a good opportunity to try a smaller startup, so I took that shot. It's still proceeding, and I have a lot of good friends there, and I have no regrets. But this guy Jonathan [Schwartz] kept calling me. You may have heard of him. There's something about working at Sun-you either love Sun or you don't. If you do, you're a Sun employee forever, whether you're working here or not.
When Jonathan called and essentially presented the opportunity of running the world's largest open-source company, and working for Jonathan-that's an opportunity that comes once in your life or less. And I had to do that-helping return Sun to where I think its place needs to be in the technology universe and fielding a vibrant open-source business.
How has your relationship with Jonathan evolved since you've returned?
Green: I think it's gotten a bit closer. We spend more time working together on more issues. Our views and strategic perspectives intersect more closely than ever. There's this process, where ultimately he's the CEO, but there are times where I lead and he follows-most times he leads and I follow. It's a remarkably smooth process.
Sun-with the release of Java, OpenSolaris, OpenOffice.org, Niagara and many of its other products to the open-source community during the last eight years-has established itself as the largest open-source IT company in the world. What else does the company need to do in the short term to close any loops still out there? Or are there any more loops to close?
Green: There are two axes of our software business. If you look at MySQL as an example, it's a model in which the vast majority of the software is made available as open source, but not all of it. Marten [Mickos, founder of MySQL and now a Sun executive] has this great expression: "There's a difference between organizations that have more time than money and organizations that have more money than time."
It's a brilliant comment. When you look at [MySQL's] stratification, all of their open-source software-which is a majority, by far, of their whole portfolio-is directed at the world community, which is dominated by folks with more time than money.
The platforms that MySQL and Sun use-which fit together so well-are there to attract developers all the way through deployment. There's no hold back in that platform that says, "If you want to go to deployment and you want to build a big company, you have to engage in a financial transaction with Sun." That would be a bit duplicitous and not fulfilling the open-source focus or business model.
But that doesn't mean that every bit of software should be open source. There are small bits of technology that are really focused on those organizations that are in full deployment but have more money than time. And so items such as optimization, configuration management, performance capabilities-those things that only are valuable when you're in large-scale deployment-are the kind of things that are available for a fee when you take a subscription for service, training and support, as well as those capabilities.
That's the overall model we're heading to. We're there in some cases but not in others. In most cases, most everything simply is open source. The incremental services are generally available as open source, as well. There are some areas in which we are moving through the open-source set of milestones, like in our infrastructure software with OpenESB and over time with our identity technology. With the big platforms-GlassFish, OpenSolaris, OpenJDK, MySQL-those are all open.
"More money than time." Is that sort of the realm of the older, more traditional companies?
Green: Oh, not at all. It's exactly the realm of medium-to-large-scale, but fast-growing, companies. If you look at one of the largest deals that MySQL has, it's with Facebook. You would not look at [Facebook] as a classic [company], by any means, but they are moving so quickly that any assistance that they can secure to maximize the use of their resources to turn around new features and services more quickly is well worth it.
That said, [Facebook founder Mark Zuckerberg] got started in his dorm room with MySQL as the basis for the platform. It is a very important point that underscores why open source and free access to technology are so important.
There is still a prevailing but incorrect wisdom that a company will start out with some basic open-source software and, as they become successful, they will step back, reassess their IT needs and reimplement on "larger-scale, more reliable enterprise deployments." So, there is the sense that propriety enterprise software will still be a successful platform because ultimately, "People will understand that's what they ultimately need."
We see existing proof after existing proof that this is simply false. The "good enough" now is absolutely great. And as long as you start out with an open-source platform that is suitable for development as well as deployment, there will never be a change. And if you try to advocate an open-source platform that is not suitable for scalable deployment, it will not be chosen.
The conceptual model of open-source platforms that are complete all the way from development to large-scale deployment is the core of the business model. There is no hold back that can be part of that business for those key large-scale deployments-or it will never realize the acceptance up-front.
MySQL has had to struggle to overturn its early reputation of being a good, little "departmental-type" database that didn't really scale or offer top-grade security. What do you tell people who cling to that idea?
Green: Well, you know, Facebook and Google are pretty big "departments." One of the value propositions we discussed before the acquisition [in January 2008] was that Sun has this worldwide service and support organization. We had the street cred to say, "We will be there for you, seven by 24, 365 [days] on a planetary scale." Frankly, as good as the MySQL team is, their size precluded them from ever being able to advocate that position.
Are they actually replacing huge propriety databases with MySQL?
Schwartz: They may not be replacing them; they're just cordoning them off, so they can contain and retire their usage. "Contain and retire"-that's the popular terminology for, "We're not going to be victimized by proprietary vendors anymore. We're moving to free software-the economic model, the innovation model and the community model." They can save a ton of money-everybody can.
How can Sun improve its market share against companies such as IBM, Hewlett-Packard and Oracle?
Schwartz: No matter how phenomenal our sales force is-and it is, precisely because CIOs want to talk to them-we're never going to win by out-hiring and out-visiting our customers. The way we're going to win is by reaching the broadest community possible and then relying upon them to introduce that innovation into the data center as time moves forward.
I was with a financial services company in which the CIO was complaining that he had to pay $40 million a year to a proprietary vendor for their database license. He said, "We would love to go to MySQL; we're just worried about its ability to scale." I said, "What's 'scaling' to you?" And he said, "We process about 2 million payments a day." I said, "Do you realize that Facebook processes about a quarter of a million transactions a second?"
There's a difference in processing a Facebook transaction and assuring a secure payment transaction, which has compliance and regulatory burdens on it.
Schwartz: There is a spectrum of applications, and the places where you require an expensive proprietary database are never going to go away. They're just never going to grow because the majority of the applications in the world aren't going to have those same "soft compliance" requirements wrapped in them.
So for the 95 percent of the other applications in your enterprise, which historically would have used that proprietary technology, now you have a wonderful alternative. And, by the way, it's an alternative that's used by some of the largest and most trusted names on the Internet today-with names like Yahoo, Google, Facebook-and globally around the world.
I think one of the things that traditionally has affected both technology companies and the media industry has been a challenge of distributing our technology across the world. Moving to free software eliminates for Sun the dependence upon traditional and proprietary distribution vehicles.
So now I can use the fact that a developer is interested in a big retailer in China to reach that developer. Just his interest is sufficient for him to look around on the Web, talk to his friends, get a positive brand image of Sun. ... And I think our brand has completely changed in the past couple or three years to be, "These guys are committed and are driving and pushing investment in open source"-and not just with their press releases, but with their R&D budget.
And I don't have to be in the same room with him or her to explain to them the value of the technology. They have immediate access to it, to the community. And, by the way, some of them-not all of them-might actually pay for the software. But all of them, bear in mind, are going to pay for the hardware. You can't download free software onto air. You have to download it into a system-systems and storage, and that's a core part of our business as well.
Will the free software giveaway promote growth fast enough to bring in the revenue?
Schwartz: Growth is the interesting question. Remember, when we acquired MySQL, they were growing at 50 percent a year. No other database company was growing 50 percent a year. The question wasn't, "Where's the growth?" The question was, "Where's the $10 billion in revenue?" Well, I can't tell you where the $10 billion is right now, but, growing 50 percent a year, I can tell you when it shows up.
On a straight-line basis, what you need to do is grow the number of new customers. If you look at the number of proprietary vendors out there, they're having a much, much harder time getting at new customers. I don't know a single startup, high-performance computing facility-I don't know a single social-networking infrastructure company that's going to rely upon proprietary software. Why would you if you have the tools of the trade available to you free of charge, and they're probably the best in the world?
Sun is back to profitability after several tough years following the dot-bomb period of 2001-2004. In your estimation, how long will it be before Sun is really back into being a true IT powerhouse-especially in the United States? Five years? Seven years?
Schwartz: The U.S. is a minority of the market. If the No. 1 application of the Internet is media consumption, the majority of the people who consume media are aged 18 to 24, and not very many of them live in the U.S., compared with the rest of the world. The U.S. may, in fact, find itself being a shrinking portion of the overall global economy from a technology consumption perspective, in part because we have relatively few people.
What is necessary to run a network of 500 million subscribers in China? We don't even have 500 million people [in the United States], much less cell phone subscribers. So the infrastructure demands are going to grow more rapidly outside the U.S. than inside the U.S.
Technically, we see much more vibrant communities-especially in the open-source world-outside the U.S. than inside the U.S. because there is still a great deal of dependence upon proprietary software in the U.S.
So when are we going to be happy with our overall performance? I'm never going to be perfectly happy. I mean, I'm not paid to be happy; I'm paid to grow the company and always be dissatisfied with our performance and to be constantly investing in our potential and realizing that potential.
I think there's been a myriad set of turnarounds [for Sun] that do not map to the last two years; they map to five or six years of R&D investment and multiple years of community investment, and we've had a fairly robust engagement from the open-source community for a while. The place where we didn't have that was around our core operating system. And that we have very, very firmly established and have driven forward.
But I don't think we're ever done. Because you just don't dump code over the wall and say, "Look, it's free!" What you do is you prove to the community that you're committed to its evolution and its investment, and that's what we're doing today with our people.
I feel better about our position today than I have in a very long time, but you have to remember that our revenue is a lagging indicator of developer adoption.
You ask me when I'll be satisfied? I'm going to be satisfied when 100 percent of the world's developers are working in and among the free and open-source software community. And I think we have a ways to go.