We are getting an error when running a Fabric Notebook with PySpark against a Data Lakehouse that has OneLake Data Access preview enabled. We are also schema enabled. The notebook attempts to run a query and it gives us an error that says:
OneSecurity error while invoking OneSecurity API isSystemContextRequired with response error {"code":"InternalError","subCode":0,"message":"An internal error occurred.","timeStamp":"2025-08-05T19:03:34.5458564Z","httpStatusCode":500,"hresult":-2147467259,"details":[{"code":"RootActivityId","message":"1e8211ef-7cda-4e89-9a52-dd9c6246d2be"},{"code":"Param1","message":"OneSecurity default error definition\n\nStatus: 400\nResponse: \n{\r\n \"code\": \"OneSecurityRoleCountLimitExceeded\",\r\n \"subCode\": 0,\r\n \"message\": \"Total role count in the GetRolesMetadata request has exceeded max allowed. Role count in request: 24, max allowed: 20.\",\r\n \"timeStamp\": \"2025-08-05T19:03:34.5209316Z\",\r\n \"httpStatusCode\": 400,\r\n \"hresult\": -2147467259,\r\n \"details\": []\r\n}"}]} and status code: 500 with error code: 500
When we researched this error, it appears there are too many nested groups for the context of the query and it’s resources for OneSecurity to work through. So, we removed several of the groups and even tested with one group on the lakehouse and it still errors and retries ~30 times. Then succeeds? We also tried creating a test user that had no groups, same result (same count of groups [24]. The normal runtime prior to this issue was only a few moments. Now it is hours. This is costing us A LOT of time and money and has impacted production workload. This HAS worked in the past and is only recently an issue since 08/01/2025. Does anyone have experience here with this issue?
