@abhyanshu1
There are different methods to accommodate this in Compose.
Option 1 - Use a LOOKUP in the ETL Mapping. The Lookup will perform a JOIN and allow you to map a column from the lookup table and bring it into the target entity in the DWH Model.
Note that when running CDC workloads - changes to the lookup table do not drive the ETL. (Only changes to the table defined as the mapping source are seen).
Option 2 - Leave the data in a normalized state in the model. A and B are managed as separate entities and use the data mart layer to denormalize the data into the dimension/fact table.
There is a whitepaper posted on community.qlik.com that discusses the modeling approach and benefits to CDC / incremental ETL and can be found here:
https://community.qlik.com/t5/Official-Support-Articles/Modeling-Real-time-Data-Warehouses-in-Compos...
Hope this helps. If need further assistance on this, please open a support case and we will get the right people to help you with this.
Thanks,
Nanda
