Issue with Upsert Table Action in Metadata-Driven Pipeline to Lakehouse Table

Hi parasol54 

Warehouse tables in Fabric Data Warehouse = Lakehouse Tables.

That means Whatever you are doing in Lakehouse ( such as creating Delta table) is also applies to Data Warehouse Tables. ( Unified Data ) 

Lakehouse Delta Tables 

Data Warehouse Tables

No limitation applies. Whether you create a Lakehouse or Warehouse Tables. I would highly recommend you to use Notebooks and Delta Tables. That way you would understand Data Lake and how it migrated to Delta Lake.

Hi parasol54 ,

Thanks for reaching out to the Microsoft fabric community forum.

lbendlin ,

jennratten ,

BhaveshPatel 

Thanks for your prompt response

Since the Copy Activity in Microsoft Fabric doesn’t currently support the 'Upsert' action for Lakehouse Delta tables, a practical workaround is to load your data into a staging table using 'Overwrite' mode, then use a notebook (SQL or Python) to perform a  merge operation into your destination table based on defined key columns this approach simulates upsert behavior by updating matching records and inserting new ones, and it integrates well into a metadata-driven pipeline if you extend your control table to include merge rules.

Lakehouse and Delta Tables - Microsoft Fabric | Microsoft Learn

Options to get data into the Lakehouse - Microsoft Fabric | Microsoft Learn

Simplifying Data Ingestion with Copy Job: Upsert to Azure SQL Database & Overwrite to Fabric Lakehouse | Microsoft Fabric Blog | Microsoft Fabric

Best Regards,

Lakshmi Narayana