Issue title Spark notebook is throwing a 403 error when calling a UDF while attached to an SPN from a Data Factory notebook activity Issue description Below cell is throwing 403 error when notebook is called from Data Factory Pipeline attached to notebook activity using Service Principal myFunctions = notebookutils.udf.getFunctions(pointer to UDF function, workspace_id) I have verify the following and they are configured properly - Add SPN/security group to the UDF workspace. - Grant Read on the UDF item. - If cross-workspace, pass the correct workspace ID in getFunctions. - Grant SPN access to the Fabric SQL DB item. - Grant SQL database-level permissions. - Confirm tenant setting Service principals can call Fabric public APIs. and still get below error from UDF artifact (udf_id) Py4JJavaError An error occurred while calling z:notebookutils.udf.getMetadataStr. : java.lang.Exception: Request to https://api.fabric.microsoft.com/metadata/artifacts/udf_id failed with status code: 403, response:, response headers: Array(Content-Length: 0, Strict-Transport-Security: max-age=31536000; includeSubDomains, X-Frame-Options: deny, X-Content-Type-Options: nosniff, Access-Control-Expose-Headers: RequestId, RequestId: 0a7ec9c0-3963-4582-be76-e369498f0ea3, Date: Sun, 10 May 2026 04:53:38 GMT) at com.microsoft.spark.notebook.workflow.client.FabricClient.getEntity(FabricClient.scala:126) Known limitations from MSFT documentation https://learn.microsoft.com/en-us/fabric/data-engineering/user-data-functions/user-data-functions-service-limits#limitations
... View more