0

We all have a scenario in our project where we are inserting data from Databricks dataframes into Azure Synapse. While we could do this without issues on Dev environment with admin access, we could not run this in higher environment. On Higher environments, Providing INSERT permission on the schema.

The error message I get…

Py4JJavaError: An error occurred while calling o2445.save. : com.databricks.spark.sqldw.SqlDWSideException: SQL DW failed to execute the JDBC query produced by the connector. Underlying SQLException(s): - com.microsoft.sqlserver.jdbc.SQLServerException: User does not have permission to perform this action. [ErrorCode = 15247] [SQLState = S0001]

piet.t
  • 11,718
  • 21
  • 43
  • 52
TANIMA DUTTA
  • 1
  • 1
  • It would help if you provides your Python code so we can validate which options/approach you are using. Different options require different permissions in Synapse – GregGalloway May 15 '20 at 02:20

1 Answers1

1

Assuming you took this approach then you will need CONTROL Database (db_owner) permissions in Synapse because it is currently required for Databricks to run CREATE DATABASE SCOPED CREDENTIAL

Though this feedback item is related to Azure Data Factory, if it were completed then more granular permissions could be used. So please vote and see my comment.

Community
  • 1
  • 1
GregGalloway
  • 11,355
  • 3
  • 16
  • 47