More relational DBMS bashing – check Dave Kellogg’s full post for more details
Mike Stonebraker spoke today at SIGMOD (see Tweetstream) where, among other things there was a 40-year anniversary celebration of the relational DBMS and, in what I suspect is non-coincidental timing, Mike did a post on the CACM site entitled The End of a DBMS Era (Might be Upon Us).
Excerpt:Moreover, the code line from all of the major vendors is quite elderly, in all cases dating from the 1980s. Hence, the major vendors sell software that is a quarter century old, and has been extended and morphed to meet today’s needs. In my opinion, these legacy systems are at the end of their useful life. They deserve to be sent to the “home for tired software.”
I have great respect for Michael Stonebraker’s accomplishments, and I am also impressed with Mark Logic’s track record, but I believe catchy phrases such as “the end of the one size fits all database” and “the home for tired software” are generally exaggerated perspectives from people who just happen to be trying to sell you non-traditional data management products…
It’s unhelpful to suggest IBM, Microsoft, and Oracle – which collectively, according to a 2009/06 Forrester report, control 88% of the (revenue share of the) commercial DBMS market – are somehow oblivious about opportunities to use new indexing and storage approaches for OLTP, leverage in-memory models for transaction processing, etc.
Mark Logic CEO Blog: Stonebraker: Send Relational DBMSs to the Home for Tired Software
2 comments:
Peter,
Apologies if redundant, but I think my first comment got lost.
In short, while you may not like Mike's slogans I have a problem with those in the analyst community that assume that just because the big-3 are big that they are unassailable. The bigger they become the more (think: Marley's chains) entangled they get with integration which definitely does impact innovation. And, with 50% operating margins and an oligopoly in DBMS, there isn't a whole lot of competitive pressure driving innovation, either, in core RDBMS land.
Technology-wise, we both know what problem the RDBMS was built to solve (flexible query to structured data). And it's had a heck of run doing so. And it's been well extended over the years to new problem classes.
But no one piece of software can be optimized for every problem. So the issue isn't black or white -- death to Oracle, long live Vertica -- it's gray. To what extent do you want information in a general-purpose DBMS vs. a special-purpose one? It is a question worth pondering because special-purpose DBMSs unsurprisingly do their "thing" incredibly fast and well.
Finally, beware the chicken/egg issue on your credibility assertions. Does Mike (or me) say what he says because he's selling a special-purpose DBMS? Or does he believe what he says and ergo decided to found (or work at) a special-purpose DBMS company?
I'm know for me it's the latter and pretty sure for Mike as well.
Post a Comment