What Red Hat missed and Sun gained in MySQL

A friend pinged me late yesterday to talk about Sun's acquisition of MySQL. In his view, "someone had to" buy MySQL, but it was a case of buying into the past, rather than the future. Application servers like JBoss are a growth opportunity, he reasoned, while the database is largely yesterday's news.

As I noted to him, this couldn't be more wrong. As Tim O'Reilly notes, with the world increasingly database-driven, owning the database (and the data it holds) is increasingly, not decreasingly, important. The database is at the heart of the web economy, and the web economy is increasingly the heart of the global economy.

It will take us awhile to get there, but the trends all point that direction.

Does this mean that Red Hat should have acquired MySQL? Probably not, for a few good reasons.

First, Red Hat's new CEO, Jim Whitehurst, believes that Red Hat needs to extract more value out of its existing markets rather than wildly expanding into new ones. As he told me, with the markets Red Hat is currently in it could be a $5 billion/year company rather than the $500 million/year company it currently is. Focus and operational excellence don't have MySQL on the cards.

Second, it's not clear that Red Hat could easily digest MySQL today. By many accounts, it hasn't done a particularly good job with its JBoss acquisition. With the dust starting to settle on that acquisition and results finally starting to churn, acquiring a new set of personalities and markets would probably be more than Red Hat could manage, especially with the added wrinkle that it also need to integrate a new CEO.

Third, it's not clear that there's significant overlap in MySQL's markets/customer demographics and Red Hat's. While this is also somewhat true of Sun and MySQL, I believe Sun has more margin for error given its size. Red Hat doesn't have the ability to take on any acquisitions that aren't immediately accretive and that fit into its sales force. I don't know that MySQL fits this bill.

A few years from now, I think Red Hat would be ready for this kind of acquisition. But not today. Perhaps it will rue the day it missed out on MySQL, but I believe Red Hat has plenty of good work to do in widening the reach of Linux and JBoss in the traditional enterprise without overly worrying about the web economy which has shown a distaste for paying for technology. By the time the Amazons and Googles of the world come around to the idea of actually paying for software Red Hat's enterprise offerings may fit a lot better than they do today.

In the meantime, Sun has done well to continue its "The Network Is the Computer" mantra with MySQL at the heart of the vision. Red Hat missed out, sure, but this just means that we have two strong open-source companies focused on taking open source into the heart of the enterprise, both online and offline.

[botchagalupe]

Mark Hinkle makes posses an interesting question?

What if Redhat were to snap up EnterpriseDB

http://socializedsoftware.com/2008/01/16/sun-buys-mysql/

johnmwillis.com

[Matt Asay]

This would require a completely different business model (as EnterpriseDB is a proprietary software company) and would be somewhat of an uphill battle as PostgreSQL doesn't have nearly the same traction/momentum as MySQL. I don't see it....

[bill-tb]

It's obvious, combining Sun's hardware data products and MySQL database is a big winner for Sun. As you state in the article, it's where the action is moving forward. Thumper, ZFS and MySQL, what a combination that will be.