Judge if the data model designed is optimized the business requirements matter most. First of all, one got to make sure the data model closely represents the business object relationship - functionally properly modeled; then considering the possible production usage situation info gathered before hands to do some load testing. Before knowing these, one just using a tool to blindly simulate data and judging if the data model design will perform well goes to the wrong direction.
In addition, your last project ran into some design and architecture issues may not only to blame the data model. If you implied it's just a performance issue, you may tune it for the production usage situation. But it is hard to tune it before you know the possible production usage situation - not necessarily wait until application is deployed and used in production. So you must do your best gather the enough information about the production usage situation before tuning it. John On Fri, Jun 19, 2009 at 11:16 AM, Edward Song <edward.s...@nuhorizons.com>wrote: > Here's a Friday discussion. > > As an application developer, I can write SQL pretty well, but I wouldn't > say > that I'm an expert. My last project I did run into some design and > architecture issues that needed a little bit of reworking and > refactoring. To help avoid this, I want to get better and have more > confidence moving forward in a project that my schema design has been > optimized, or at least perform some excercise to let me know that I gave it > a good effort. > > Any insights on how as application developers, we can perform some exercise > with some tool at the early design phases that will help us give us > confidence that our initial schema, indexes and queries are optimized? >
