As mentioned in my previous post, we are looking at a different way to produce the metadata needed for the Data Vault Framework. Excel isn’t our thing.
We are currently using Oracle for our database and do the modeling of the Data Vault by hand using Oracle SQL Developer Data Modeler. We believe that modeling by hand is better than a Data Vault model that is generated based on a source model.
In the modeler, we are able to specify the metadata we need for the automation framework. We use the “notes” field of tables and columns to specify the source(s).
Getting that metadata out again in a format we want, is quite a challenge. Up to now, we have three different possibilities:
- use the reporting schema capabilities, which basically stores the model into a database schema with a lot of tables, and write SQL queries to combine the necessary information and load it directly into the metadata tables that are used by the automation framework
- use transformation scripts within the modeler, written using java script, that take parts from the underlying design and “does something with it (still to be defined what exactly)
- use a combination of both
To be continued…