Since its spinoff, Ingres has made no bones about setting its cap on stealing Oracle customers. This is no MySQL; this is no company that demurs and calls itself a commodity database. This is a company thats out for pure, red, enterprise database lifeblood.
Still, as a company, Ingres is 90 days old Feb. 7, but we havent heard much about its business plan or what that plan will mean to the open-source community.
As it is, its already got a management team thats pure Redwood Shores: former Oracle Senior Vice President of worldwide marketing and business development Terry Garnett is now chairman and CEO; former support and maintenance exec Mike Rocha is a board member; former applications development exec Mark Barrenechea is now a CA board rep; and former “Unbreakable Linux” exec Dave Dargo is now Ingres CTO.
As for Ingres newly appointed CFO, Tom Berquist, hes still a Citigroup employee until March 1.
As managing director, hes still covering companies such as Oracle, Microsoft and BEA. He cant talk dirt about Ingres plans to undercut, say, up to 8 percent of Oracles database customer base.
Not specifically, that is. He managed to do so implicitly, though, in a recent conversation with Database Editor Lisa Vaas, along with sketching his business plan for Ingres and how the company plans to work with and foster the open-source community around the database.
Youve been watching open-source companies get commercialized since you were an analyst at Goldman Sachs and Red Hat was coming together. What are you going to bring to Ingres development from what youve observed?
As the Internet bubble collapsed, a lot of people left open source for dead and went to more traditional vendors. But we noticed, in the early 00s, the big challenge to corporations was that IT budgets had been slashed. They couldnt afford to pay the prices for technology they had in the 90s.
We researched and found there was opportunity for open-source players to sell services around open-source products that would be dramatically cheaper than their full-priced competitors.
In the cost of building a software company, probably 35 percent is attributable to R&D, and the rest is sales and marketing and going GA and the rest of the stuff.
The open-source model, with a community of developers, can significantly cut back on some 35 to 40 percent of the costs of development.
Further, the community contributes ideas to take the product in areas you otherwise wouldnt have come across.
Of course, the negative is that dollars coming in on a per-unit basis are lower, so you need more users paying a small amount, vs. the enterprise model of few users paying a large amount.
The opportunity we see is to take enterprise mission-critical technology used by 10,000 customers over the last 25 years, take it to a very large number of companies, both [Ingres] existing base and to new targets who are trying to build out new Web-based applications or portals, maybe Web 2.0 technology, and give them a cheaper deployment infrastructure that would hopefully enable then to go out and do even more projects they otherwise wouldnt be able to do, because the original project would be so expensive it would pre-empt their ability to do other things.
So Ingres business model supposes that a large enterprise can shift off of large enterprise systems and safely go to an open-source infrastructure, in part or in full. Any examples of that being done without the company going kablooey?
Sabre, the travel reservation processing company, they used to run on all commercial infrastructure. The cost was gigantic. It was trying to roll out Web services for [new capabilities].
It rewrote its architecture to use a large percentage of open source. It reduced cost of deployment dramatically, because it turns out an awful lot of reservation processing is you deciding you want to fly from San Francisco to Boston … and you want to see flight options.
You might check three destinations, maybe a few dates, and if the site is lucky, you book through them. They cant afford to use big, expensive [infrastructure] for every transaction, so they used open source.
The argument Im making is every company has this opportunity.
Every company. Which means Ingres is setting its sights on a broad array of industries, such as …?
Telecom, Internet companies, financial services companies, large multinational manufacturers, retailers. Big heavy IT users. Ones with massive data centers, often distributed. Some may run 30,000, 40,000 or 50,000 database instances.
Were not trying to replace all [those enterprise database instances] with Ingres but would love to do that over the course of time. In the near term [were aiming to get] 5, 6, maybe 8 percent penetration by helping with applications.
Eight percent of the market is a lot of money. What makes you so sure Ingres can talk that many people out of their proprietary enterprise databases?
Its a $500 million to $700 million market were talking about. Youd get a pretty decent chunk. Now, there are plenty of competitors for that chunk. There are plenty of people looking at ways to extract some of that. We just have a better database product.
Its enterprise scalability, its customer list, its been proven over a long time, its good performance, the fact that we own all the technology, and we can fully identify customers, where some [open-source database companies] dont own all of the technology so its harder for them to do that.
We believe we have pieces that enterprises would need if they took a piece of their infrastructure and put it on an alternate provider.
What are you doing to build community around Ingres or to tap into what the existing community has already coded up?
There are two classes of community: the customer community whos been using Ingres for many years. Our goal is to help marshal some of those organizations into contributing things theyve written for the Ingres database, to contribute through the open-source community, back to us.
What specifically are the features youve seen out there that youd like to get into the main code set?
Theres definitely been a lot of people whove OEMed the product. When OEMed into a third-party application, you try to significantly reduce administration around the product. One thing weve heard from people whove embedded our database is that it requires remarkably little administration from people whove embedded it.
The rules and processes people have developed to keep it available without the care and feeding youd expect.
One quote a customer made was he doesnt understand how companies can afford [database administrators].
What about the developers whove been using Ingres since CA open-sourced it in 2004?
We know theres a community developing around it. Theres a user group and others whove developed [around the database]. We know there are current initiatives there as well.
The goal is to marshal the community and have a process of submission of code back [to Ingres]. For example, we know people have written converters for other database vendors products to our product. That was the origin of the $1 million challenge.
Editors Note: This story was updated to correct Terry Garnetts title at Oracle.
Check out eWEEK.coms for the latest database news, reviews and analysis.