I picked up via this post on the High Scalability Blog a new paper by Michael Stonebraker, Nabil Hachem, and Pat Helland entitled The End of an Era (It’s Time for a Complete Rewrite) presented at the VLDB 2007 conference in Austria on September 23rd through 27th.
From the paper’s summary:
“In the last quarter of a century, there has been a dramatic shift in:
1. DBMS markets: from business data processing to a collection of markets, with varying requirements
2. Necessary features: new requirements include shared nothing support and high availability
3. Technology: large main memories, the possibility of hot standbys, and the web change most everythingThe result is:
1. The predicted demise of “one size fits all”
2. The inappropriateness of current relational implementations for any segment of the market
3. The necessity of rethinking both data models and query languages for the specialized engines, which we expect to be dominant in the various vertical markets”
As you know, I’m a big believer in the special-purpose DBMS meme. Any database historian knows what Codd was thinking, and more importantly — what he wasn’t — when he designed the relational model. Again, from the paper:
“Ted Codd’s idea of normalizing data into flat tables has served our community well over the subsequent 30 years. However, there are now other markets, whose needs must be considered. These include data warehouses, web-oriented search, real-time analytics, and semi-structured data markets.”
The complete paper can be found here.
