We at Headspring uses Tarantino for database change management. Earlier I have blog it’s ability to skip certain script file and how to use this feature.
As per the standard Headspring practice, we develop our project in the environment that facilitate close collaboration with customer. Which implies that our customer is seating next to developer team and will test user story as and when it gets completed by development team. This unique setup enable early feedback to hone the end product. Since customers are continuously testing system, they need to maintain right set of data that represent real life scenario. While testing specific scenario, customer will enter some supplemental data and perform transactions which again ends up modifying core data. This obviates need of resetting database to known state to perform further testing.
With Tarantino, deploying your application to staging environment is automatic. (If you aren’t aware of this feature of Tarantino, go and read it from here.) Hence if can capture script (both schema and data) that represent current version of database at any given time, then later on resetting database back to known state is simply mean recreate your database from this script.
Catering to this need, I have put together utility which will generate schema and data script for given database instance (Works only with SQL Server, any version as it uses SMO library which is specific to SQL Server)
We have been using this utility for over more than month and it make our database resetting step easy enough for us to forget about that step at all.
If you are curious and have similar need, you can grab the utility from the following SVN repository. You need svn client to download it. If you haven’t got it already, get one from here.
http://svn2.xp-dev.com/svn/mdmavani-DbSnapTaker/