We are just starting with Pentaho DI and don't want to re-invent the wheel.
We are looking for advice on what would be the best way to use Spoon / Kettle facilities to perform the typical Dimension / Fact building activity? Is there a standard / best practice way to do it? Spoon Steps designated to perform it? Any tips on how to potentially do the lookups/inserts of all Dimension tables in parallel?
1. The process reads next detail (Fact) record.
2. Some attributes of this Fact record need to be moved into Dimension tables and replaced with the id of the corresponding record in the Dimension table.
3. In order to get this id, we search the Dimension table using one or more Detail table attributes.
3.1. If a corresponding record is not found, it is created using all the available search attribute(s).
3.2. The id of the found or newly created record is passed back and assigned to the column in the Fact table.
Thank you,
-- Alex
We are looking for advice on what would be the best way to use Spoon / Kettle facilities to perform the typical Dimension / Fact building activity? Is there a standard / best practice way to do it? Spoon Steps designated to perform it? Any tips on how to potentially do the lookups/inserts of all Dimension tables in parallel?
1. The process reads next detail (Fact) record.
2. Some attributes of this Fact record need to be moved into Dimension tables and replaced with the id of the corresponding record in the Dimension table.
3. In order to get this id, we search the Dimension table using one or more Detail table attributes.
3.1. If a corresponding record is not found, it is created using all the available search attribute(s).
3.2. The id of the found or newly created record is passed back and assigned to the column in the Fact table.
Thank you,
-- Alex