Bye-bye, Big Red? Escaping Oracle’s not that easy

Talk to certain folks and you’d believe Oracle is doomed — doomed! Its proprietary database solutions are a prime target for being displaced by the likes of open source contenders MariaDB and PostgreSQL. It’s only a matter of time.

Sure, but the real question remains: How much time? And what will it take to make leaving Big Red behind a real proposition for an enterprise that has Java middleware, Sun servers, and Unbreakable Linux running all of the above?

The folks behind PostgreSQL and MariaDB have a lot to crow about. PostgreSQL is widely respected as an enterprise-class database, even if it doesn’t get the kind of press MySQL routinely did. One of the biggest PostgreSQL providers, EnterpriseDB, recently landed a plum spot in the Challenger Quadrant for Gartner’s Magic Quadrant for Operational Database Management Systems report. I’m just as skeptical as anyone else about the value of the Magic Quadrant reports, but there’s little doubt that being featured there gets you the right kind of attention if you’re a growing company or dark horse.

MariaDB, the MySQL fork, has also been flush with successes — both strategic and financial. The company has made enough of a dent in the consciousness of the digerati that many longtime MySQL users are switching to MariaDB: Wikipedia and Google, just to name two. Given Oracle’s shabby track record with MySQL and unfriendly pricing, it’s hardly surprising.

MariaDB’s also inspired others to take MySQL and beat Oracle at its own game. To wit: Percona, who has its own edition of MySQL 5.6 that sports variants on features found in Oracle’s enterprise editions of that product. The MariaDB folks have also benefited from a $20 million infusion of cash courtesy of SkySQL.

This is all great news, but let’s not confuse any of this with a direct assault on Oracle’s customer base or business plan. These things by themselves don’t constitute displacement of Oracle, which has a staggering $37 billion revenue base as of 2013 and has even modestly goosed its profit margins across the past five years. Expecting Big Red to dry up and blow away any time soon is foolish.

Because so much of the press about Oracle revolves around its products and technologies, it’s easy to forget that Oracle is primarily a services company — one that happens to be in the market of selling its proprietary products to Fortune 100 customers. Folks like PostgreSQL and MariaDB may not be able to tear those customers away from Oracle all at once — not just because Oracle makes it difficult to leave, but because these customers are invested in Oracle as an application provider, not just a database provider.

That migration can happen in time, but it’ll take a lot more than infusions of cash and a roster of success stories from companies that would never have used Oracle in the first place. If PostgreSQL and MariaDB are serious about moving people off Oracle’s platforms, plural, they’re going to need to provide all the things Oracle does: databases, apps, and support structure. The first and third, they already do; there’s plenty of outfits from which to buy enterprise-level support for both competitors’ products or which supply migration tools.

But once people have a taste of how thorough Oracle’s product support is, they’re reluctant to let go. Oracle’s as hard to get rid of as a legacy mainframe system. Despite all the crowing that went on about Java being dead as an enterprise app platform, it remains firmly entrenched, and Oracle plans to make the most of it via everything from the Internet of things to the various surprising uses of the JVM.

The real place where PostgreSQL and MariaDB can and have made inroads is with everyone who comes in with a clean slate. When two guys in a garage want to start something big, they spin up some Amazon EC2 instances and go to town, and make MySQL (or MariaDB) part of the plan. When a company like that hits it big, it turns to the likes of Salesforce or SugarCRM for its line-of-business app needs. No greenfield projects use Oracle — not when there are plenty of ways to build great things from scratch without the merciless lock-in of Oracle.

What would a real Oracle killer look like? It wouldn’t be any one database solution, that’s for sure. It would be more like a whole migration services platform — one where you could have experts come in, look at your Oracle application and DB mix, and construct a custom migration solution. That’s not something you can provide as a download; it has to be specially put together.

It’s reminiscent of the way people used to talk of Linux being an across-the-board Microsoft killer. Today, Linux is ubiquitous and indispensable, but Microsoft is still very much alive and kicking — the black eye of Windows 8 notwithstanding. Some of that is embrace-and-extend on the server side: It’s done its best to make Windows Server run Linux well inside Hyper-V, and it’s bolstered support for many common open source frameworks and languages in Windows Server. But few people expect the desktop world, where Microsoft reigned supreme for so long, to be its central market for much longer.

A lot of that is tenacity of a sort that Oracle itself is likely to exhibit — that is, once its business really starts getting eaten into on the legacy high end by the likes of all that open source competition.

This story, “Bye-bye, Big Red? Escaping Oracle’s not that easy,” was originally published at InfoWorld.com. Get the first word on what the important tech news really means with the InfoWorld Tech Watch blog. For the latest developments in business technology news, follow InfoWorld.com on Twitter.