I'm interested in playing around with the EAV model over an SQL Server Database. Is anybody able to point me to an article or framework for an existing model/schema (I'm not intrested in Saas for this)?
Best Practices for Semantic Data Modeling for Performance and Scalability
EAV is notoriusly problematic as it leads to severe deployment performance and scalability problems. The Whitepaper in the link, released by the SQL Server Customer Advisory Team tries to offer some guidance to deploy a succesful EAV model.
Why not create a very simple table with an entity column, an attribute column, and a value column:
CREATE TABLE eav(
entity STRING,
attribute STRING,
value STRING);
And then populate the table using a script of some sort to generate some test data.
INSERT INTO eav ('banana','color','yellow');
INSERT INTO eav ('banana','fluffy','no');
INSERT INTO eav ('banana','origin','Guatemala');
INSERT INTO eav ('orange','origin','USA');
INSERT INTO eav ('orange','origin','Mexico');
INSERT INTO eav ('pear','origin','USA');
INSERT INTO eav ('peach','fluffy','yes');
And then try running queries on it, to play around, learn how it works, etc.
In general though, the above data organization scheme doesn't really mesh well with relational database theory. The above meshes more with document database theory, like Couch DB. I'd look more into that if you need to store/manage data of EAV nature in the wild.