ISVs and Oracle: A Troubled Tango
There are more defections from Oracle than people expect, with many ISVs looking for economic savings or an easier upgrade path, Yong claims: "Those who move from Oracle 7 and 8 to 9i made a huge leap—and it would have taken the same amount of effort to move to SQL Server," Yong says.

Indeed, most of Scalability Experts clients have a mix of installed database software, usually one version behind the latest release. While end users may have a broad mix, ISVs try to reign in the maintenance complexity by restricting themselves to two or three versions.

"ISVs tend to build vertically focused software solutions such as telecom or healthcare," says Microsoft Alliance Director Derick Schaefer. "In many cases, 20 years ago they went down the Oracle path. But as customer needs change, or as they consider completely re-architecting their solution, these ISVs see that today, SQL Server does good job of taking care of itself, especially in small-to-medium size businesses."

Also attractive to leaner budgets is the fact that SQL Server is a great choice for the business intelligence space, with accessible and economical BI capabilities. Say you're an ISV that writes inventory apps; with SQL Server 2005, you can offer your customer a reporting system with different data views and analysis to figure out run rates and seasonal stock levels. With Oracle, Yong argues, "You'd pay tens of thousands of dollars in many cases for their analysis option."

Total Cost of Administration
"We've seen SQL Server really lower our cost of administration," according to Yong. A decade ago, version 7.0 took away a lot of the routine management tasks such as configuring memory and CPU behavior manually. For example, SQL Server automatically determines how to dynamically allocate memory to different caches or automatically determine if parallelism was appropriate for a query, all managed at the engine level without user intervention. "You had to know what you were doing to write queries that ran well; if your workload or system resources changed, you had to change either your queries or the system configuration (e.g. switching from OLTP to DSS type queries)," he explains. In response at the time, Larry Ellison jibed that Microsoft didn't understand databases. Today, the tables have turned, with Oracle preaching self-managing databases.

And really, which sounds more fun: working on a team of DBAs who continually stare at performance counters and occasionally tweak them, or working with architects and developers who turn state-of-the-art business intelligence tools into novel, customer-pleasing solutions?

In the end, Oracle continues to aim at the enterprise space, while Microsoft starts at bottom, with small business applications, and goes all the way up to SQL Server implementations supporting thousands of concurrent users and tens of Terabytes of data. As data becomes more ubiquitous and decision-support more mission-critical, Microsoft's migration strategy just might pay off for ISVs wanting to maximize money—and great minds.

* This article was commissioned by and prepared for Microsoft Corporation. This document is for informational purposes only. MICROSOFT MAKES NO WARRANTIES, EXPRESS OR IMPLIED, IN THIS SUMMARY.