The relationship between metadata, data source and data warehouse is critical to understanding how Ajilius works.
We use the term metadata to describe the tables, columns, calculations, documentation and other information that helps us to build the data warehouse.
We take metadata from source systems (tables, columns and data types), store that metadata in the Ajilius repository (with documentation, calculations, etc.), then use that metadata to create the tables in your data warehouse and the scripts that load and transform those tables.
<< diagram >>
When you decide that you want a table loaded to your data warehouse, the first step is to bring its metadata to the Ajilius repository. You do this by selecting the Load Metadata option from the data source table’s context menu. Here, we are going to load metadata for the Employee table in the ChinookOLTP data source:
Load table definition
The next screen supports entry of the information about the table, as you want it to be named and described in your data warehouse.
Refer to the documentation about editing tables for more help with using this form.
When you have saved your metadata changes, you will see a screen like the following image. The new metadata has been added to the Ajilius repository, and we’ve highlighted the table and the columns for which metadata has been imported.
You should now review the section on managing columns, and define business keys and other column-level metadata related to your table. The earlier this is done in the design process, the more rich and complete your documentation will be.