Forum Discussion
Microsoft ODBC Driver for Data Engineering Failed to Create Spark Session
Hi folks,
I have been trying the newly released Microsoft ODBC Driver for Microsoft Fabric Data Engineering - Microsoft Fabric | Microsoft Learn. I came across an puzzling error after setting everything up:
Here is where I am:
- Windows 11 64bit 25H2
- Have access to workspace, lakehouse as "admin" role
- Fabric ODBC driver 1.0.0 with DSN setup, connection tested successful
- VSCode, Python 3.14.2, using pyodbc module, using "Interactive" authentication, authenitication also successfully completed
What other settings that I should be looking at from here?
Thank you all very much.
Andrew
Hi Andy_JH ,
Thank you for reaching out to the Microsoft Community Forum.
The error indicates, the ODBC driver can’t locate one of the Fabric objects it needs to open a Spark session, because a GUID (WorkspaceId/LakehouseId) or region/endpoint in the DSN/connection string is wrong or mismatched.
Please try below things to fix the issue.
1. Please check WorkspaceId and LakehouseId the GUID is correct or not.
2. Please check fabric workspace’s region and the DSN’s configured endpoint/region must match. Ensure you are logging into the same Entra tenant that owns the workspace when using Interactive auth.
3. Please check you have access to the specific Lakehouse item, if item-level permissions are used.
4. Check your ODBC driver, TDS/SQL endpoint (ODBC Driver 18/SQL Server), used for Warehouse and Lakehouse SQL analytics endpoint. Spark/Livy ODBC, used to run Spark SQL via Livy against a Lakehouse (Data Engineering).
5. Delete the old DSN and re-enter values. minor naming/format variations can cause issues.
I hope this information helps. Please do let us know if you have any further queries.
Regards,
Dinesh
3 Replies
Hi Andy_JH ,
Thank you for reaching out to the Microsoft Community Forum.
The error indicates, the ODBC driver can’t locate one of the Fabric objects it needs to open a Spark session, because a GUID (WorkspaceId/LakehouseId) or region/endpoint in the DSN/connection string is wrong or mismatched.
Please try below things to fix the issue.
1. Please check WorkspaceId and LakehouseId the GUID is correct or not.
2. Please check fabric workspace’s region and the DSN’s configured endpoint/region must match. Ensure you are logging into the same Entra tenant that owns the workspace when using Interactive auth.
3. Please check you have access to the specific Lakehouse item, if item-level permissions are used.
4. Check your ODBC driver, TDS/SQL endpoint (ODBC Driver 18/SQL Server), used for Warehouse and Lakehouse SQL analytics endpoint. Spark/Livy ODBC, used to run Spark SQL via Livy against a Lakehouse (Data Engineering).
5. Delete the old DSN and re-enter values. minor naming/format variations can cause issues.
I hope this information helps. Please do let us know if you have any further queries.
Regards,
Dinesh
Hi Andy_JH ,
We haven’t heard from you on the last response and was just checking back to see if you have a resolution yet. And, if you have any further query do let us know.
Regards,
Dinesh
Hi @Andy_JH ,
We haven’t heard from you on the last response and was just checking back to see if you have a resolution yet. And, if you have any further query do let us know.
Regards,
Dinesh