I rarely write a column that that has more to do with Oracle than SQL Server. In fact, I think this is the first time. However, I recently ran across an intriguing open-source RDBMS called EnterpriseDB Advanced Server that’s based on the PostgreSQL database. The most interesting aspect of this database, at least to me, is that it claims the ability to fully automate the conversion of an Oracle application to EnterpriseDB. Not “We’ll get it really close and then you do the rest of the work,” which is what most database conversion tools do. The EnterpriseDB Web site ( http://www.enterprisedb.com/index.do ) actually says “After the data and business logic has been transferred from the Oracle database to the EnterpriseDB database, the database application originally written for the Oracle database may be run on the EnterpriseDB database without any re-coding or translation.”

It seems that the native procedural language in PosgreSQL is called PL/pgSQL and that this language is designed in a way that allows for a full and complete migration from PL/SQL. A press release from EnterpriseDB ( http://www.enterprisedb.com/news_events/press_releases/15_08_06a.do ) offers more information about the newest version of their flagship product, which offers the automated migration capability. You can read about how EnterpriseDB’s pricing and support models work at http://www.enterprisedb.com/products/licensing_pricing.do .

A quick review of the pricing model shows that the product is free, and you can acquire various levels of service and support that range from “still free” to $1000, $3000, or $5000 per CPU, depending on the level of support that you want. (Note that EnterpriseDB defines a CPU at the socket level, so a dual-core CPU counts as a single CPU.) Needless to say, that’s significantly less expensive than Oracle, and oh yeah--did I mention that the Replication Server product can allow the migration from EnterpriseDB back to Oracle again, supposedly in an automated, turnkey manner?

Personally, I’ve never thought that open-source databases had much chance of securing a large piece of the enterprise database market. I based my opinions on many reasons that I don’t want to defend or elaborate on right now. Instead, I wanted to think out loud about whether EnterpriseDB could change the way enterprise database customers experiment with open-source databases.

Consider the following scenario:

  • You work for BigGiantCompany and spend substantial sums of money on your Oracle investment every year.
  • You (like me) have a hard time seeing open-source databases ever being up to the challenge of replacing your core database investment.
  • You’re awfully busy and don’t have time to dabble with a long project to see if open source might work for you.
  • But, what if you found an open-source test that you could use for free in days or a week? In addition, the test uses a fully automated conversion tool that allows your existing applications to run with no change? And what if you could download that new database for free?
  • Let’s assume that you (a stubborn DBA, set in your ways like many of us are) still don’t want to test it? What happens when your boss, or your boss’s boss, says, “Hey, this should only take a few days to test and we’re going to do it before writing another seven-figure check to Oracle?”
  • Hmmm. Could that scenario open the floodgates to open-source experimentation and dabbling among Oracle-based shops that would otherwise have no interest or time to experiment and dabble with open source? I suspect it would. I also suspect that SQL Server, IBM, and other proprietary database shops would be more inclined to experiment and dabble with open source if they started reading lots of news stories about Oracle shops reaping huge savings. I bet the CFOs of those organizations would be interested, even if the DBAs weren’t. EnterpriseDB should be an interesting product to watch as it matures in the enterprise database market.