MySQL AB is making the point this fall that its open source database is enterprise-ready. The company ended October with the announcement of a huge project with Sabre Holdings and on Nov. 17 began shipping a heavy-duty, SAP-certified database called MaxDB. In this interview, MySQL CEO MÅrten Mickos explains why the IT industry is surprised at his company's enterprise success and tells where MySQL is going from here. MySQL AB is based in Uppsala, Sweden, with U.S. headquarters in Seattle.
How will MaxDB change MySQL's road map?
MÅrten Mickos: MaxDB, which we inherited from SAP and is SAP-certified, can run an SAP R/3 installation. R/3 is perhaps the most demanding ERP application you can find in the world, and we have a database that can support it.
We've had a road map for how to develop MySQL, and we haven't changed that road map. But think of it like this: We were sitting in a rowboat rowing as fast as we could. Then SAP comes up and says, 'Hey, guys, here are two Mercury 100hp engines that you can use to get to your goal.' So, SAP has had a lot of impact on our road map. We see the milestones passing faster than before.
Some of our readers expressed surprise that MySQL was being used for a huge enterprise project like Sabre Holdings' fare-lookup application. Was that an unusually big project for MySQL?
Mickos: This is what we do everyday. This is just one of the large projects we've done -- like Cox Communications, [the] Associated Press, and the state of Rhode Island -- that we have published to the public.
Of course, the Sabre project is fantastic. It's a Web application where Sabre is using it, but it's a mission-critical, heavy-load application. They have been playing around with MySQL for a long time. They had already stress-tested the database on the volumes and the responsiveness, so they knew what they were getting into. They also knew that they could depend on MySQL as a company.
Why do you think people were surprised that MySQL is being deployed at an enterprise level?
Mickos: It's true that our legacy was to be a lean, fast, fairly simple database. That's where we come from. So, I can't blame some IT managers who got the impression a few years ago that MySQL was limited to low-end, low-volume situations.
Three years ago, we were surprised at the interest expressed in MySQL. We hadn't expected that conservative, cautious corporations would be so ready to expand their open source usage from the operating system to the database. It is happening faster than we anticipated.
When the corporations first started coming to us, we didn't really have all the [manpower] to support them. A year ago, we started wrapping up partnerships and building sales and support teams. We are ready to support enterprise customers.
Now, the rate at which we get into new projects is growing.
How did MySQL attract corporations' interest in the first place?
Mickos: Sabre is a great example, because it shows how MySQL penetrates and then radiates. They used MySQL for one Web project. Then, they say, 'Can we do something else with you? You've proven yourself.'
We penetrate from the Web or departmental application and then start radiating. The rumor spreads that some part of the company is using a really fast affordable database that never has crashed. Then they come visiting the bright guys who did the first implementation and then it [starts] building from there. Then we start radiating within the organization.
It gives me a kick to see that customers are not just open to our messaging; they're standing there waiting for us when we knock on the door.
Isn't it hard to compete with major players like Oracle?
Mickos: We have no conflict with Oracle. Our most successful customers typically run some of their applications on Oracle and other parts or their database structure on MySQL. We are complementing the offerings of the huge enterprise database vendors.
When is MySQL not a good fit?
Mickos: There will always be situations where we are not a good fit, and we don't mind admitting it. Customers sometimes see only black and white. Either your product is good and will run everything, or it's bad for everything. MySQL is a great fit in the growing segments of new Web-based and Web-related applications and departmental applications.
That's not to say that MySQL won't stretch beyond those boundaries. There's an underlying thing here that isn't obvious. Traditionally, it took 20 to 30 years to build a proper enterprise database. Now, we come in and do it in only 10 years. People say it can't be done because look at how long it took to build Oracle. People are not ready to accept the fact that software technology and software models have changed.
Databases are becoming commodities, which means that you can quickly go from concept to completion. We still have MySQL Classic, which is an easy, simple, but very fast and reliable Web database. That same source code added with some other modules is now an enterprise-ready database. That's shocking news for some people.
What's ahead for MySQL technologies?
Mickos: We have three main product priorities: performance, reliability, ease of use. Within these priorities, we are moving ahead of implementing all mainstream database features -- stored procedures, triggers, views, and other no-brainers -- that many other companies implemented earlier to achieve feature parity. Looking at new developments in areas like Web services, we are developing better replication, synchronization and new clustering technologies that make it possible to assemble vast databases out of cheap hardware and software components. We're out innovating and creating new stuff, thinking about what the Web-enabled world of tomorrow really needs.
We support mission-critical databases very well from the transactional standpoint. But on the tactical side, we need to make it easier to build and port applications to MySQL. So, stored procedures, triggers and views don't really change the reliability or business power of the database, but they do make it easier to build a business application. The transactional side will always be the core of database computing that handles accounting and other applications without losing a line of data.
FOR MORE INFORMATION:
FEEDBACK: Is your enterprise using MySQL as a mission-critical database?
Send your feedback to the SearchEnterpriseLinux.com news team.