Over the past week, I've written a few entries regarding complex databases with interesting characteristics for testing purposes. 

• Creating Complex Test Databases - Intro 
• Creating Complex Test Databases - Creating a Database with 1 Billion Random Rows 
• Creating Complex Test Databases - Generating One Table for each of the 2,397 supported Collations 
• Creating Complex Test Databases - Creating a Table with all Supported SQL Server Data Types
• Creating Complex Test Databases - Creating a Table with more than 1,024 Columns
• Creating Complex Test Databases - Using Excel to convert a raw list of text into viable T-SQL syntax

I selected these first topics loosely on the fact that I was going through the exercise of creating these specific databases for testing purposes as part of my "real" job.  Now I would like to turn the question to everyone else to drum up some more interesting ideas.

Are there any complex/interesting databases that you'd like to see highlighted in a blog entry? 

A few thoughts come to mind surrounding:

• Complex relationships of PK/FKs, including composite keys
• Interesting user/login/schema relationships
• Using column sets to exceed the 1,024 column limit in tables
• Sql_variant data of all supported types  (Ex:  Sql_Variant data type can store data in the form of 23 other system data types.  If your application contains any sql_variant columns, testing that the app works with data of all supported types is a good practice.)
• Exploring UDTs and UDDTs
• Spatial types and data
• Circular object dependencies
• Stored procedures / Functions that refer to objects that no longer exist

I will eventually compile all of the scripts and documentation from this blog series to publish in the MSDN Code Gallery as a download.  Leave me a comment with any ideas that you would like to see included in the sample code.

Hope you enjoy,
Sam Lester (MSFT)