You can use the data hub to normalise incoming data, and to split a de-normalised data source into multiple tables.
This follows the same rules as any other mapping, but can seem difficult to follow.
The key points to bear in mind are:
- All the tables that you are trying to write to should have the same source entity reference. Use the source reference override on the source system or entity mappings for these.
- When you are adding the source entities to the source system, put them in the order in which they should populated. This means parents should come before children.
- If you are also linking to the normalised tables (which is probably what you'll want to do), you will separately need to include entities or mappings for these to act as the link entities for the link fields. You can use the source reference override on the source system definition to refer to the same source entity twice.