>ABAP has the referential integrity logic, locking, number range generation, etc. For these actions you should call back to ABAP for the Create/Update/Delete operations.
Would the recommendation/guideline for any custom development on ABAP on HANA to keep all Insert/Update/Delete operations in the ABAP layer? HANA native development can obviously implement these but is it best to avoid having to write these from scratch at the database layer?
I did ask this originally around having a HANA XS app having dml access to say for example Z tables in SAP<SID> schema, would this currently not be recommended/supported?