Allan Parker of Sun has written a fascinating post on the rising challenge that open source database products (OSDB: such as MySQL) pose to proprietary products such as Oracle and DB2. This post is a must-read for all database administrators.
Update November 11, 2008
As commenter Leons Petrazickis notes, Sun has acquired MySQL. In fact, Mr. Parker’s employer acquired MySQL two months after his piece was posted. In his prophetic article Parker predicts that one of the ways that proprietary vendors such as Oracle or DB2 will stave off competition by OSDB’s is through acquisition of the products themselves. In a scathing critique of the purchase, another writer John Dvorak suggests that the only beneficiary from the purchase is Oracle, a Sun strategic partner.
Clearly the struggle by the big database vendors to maintain market share relative to the OSDB’s has begun. Parker makes a persuasive argument that processor-based pricing by IBM and Oracle has resulted in increasing Total Cost of Ownership (TCO) for their databases, especially compared with the declining cost of the hardware that the database resides on. Even though the cost of ownership continues to go up, DB2 and Oracle features have stagnated and OSDB benchmarks relative to DB2 and Oracle continue to improve. OSDB’s look increasingly attractive, especially to new companies, offshore companies, and innovative new technology firms such as Google, Facebook, Youtube and Flickr.
Commenters counter that change in corporate database culture occurs very slowly. We will all have a few years before we need to worry about declining market share for DB2 and Oracle. Not only do database products change slowly, so also do internal corporate cultures (which condition attitudes regarding acceptance of new products) and the availability of trained employees and contractors. One commenter, for example, points out the difficulty of finding experienced PostGreSQL developers and administrators. And even though OSDB’s are making progress in both performance and features, this also is glacial, and equality with DB2/Oracle is not imminent.
I believe, however, that database administrators and developers always have to keep their eye on the Next Big Thing, and not content ourselves with the promise of slow change. The open source revolution will eventually affect us all. What, in particular, can DB2 database administrators and developers do to remain relevant in the IT marketplace?
- Parker points out that both IBM and Oracle can slow down the advance of the OSDB’s by maintaining a more competitive pricing structure. Both DB2 and Oracle are too expensive. Database administrators and developers have a vested interest in keeping DB2 in the market; we need to be activists on the DB2 pricing issue. Don’t be afraid to let IBM know that we don’t want DB2 riding off into the sunset. IBM, with its diverse revenue base, is better positioned than Oracle to reduce the price of its flagship database product.
- IBM can continue to position its freeware DB2 Express-C as a viable OSDB competitor, gradually augmenting its memory and processor limits to match conventional commercial usage. This allows cash-strapped companies, especially in India and China, to consider DB2 for mid-range databases.1
- Relational database administrators should continue to be open to opportunities to crosstrain in other database systems, particularly PostGreSQL. Many DB2 database administrators, myself included, have looked across the fence at Oracle and considered cross-training, or even a complete switch to Oracle. Parker shows that DB2 and Oracle may both, eventually, face extinction if they don’t make big changes. Maybe DBA’s looking for job security should look in another direction altogether – toward the OSDB’s.