Authentication error when using imported data asset from SQL in AutoML job (Azure ML Studio)

Question

I'm running into an authentication error when trying to use an imported data asset in an AutoML job within Azure ML Studio. The error message I'm receiving is:

ScriptExecution.StreamAccess.Authentication

PermissionDenied(Some(The authentication information was not provided in the correct format. Verify the value of Authorization header.))

Here are the steps I've taken so far:

1. Azure SQL Server setup:
   • Public network access is set to "Selected networks".
   • Firewall rules are configured to include our local client IP and two IPs for the Azure ML Studio resource (20.41.3.29 and 20.41.2.48).
   • "Allow Azure services and resources to access this server" is checked.
   • SQL server connection policy is set to "Default".
   • No outbound networking restrictions are in place.
2. Create an Azure SQL DB connection:
   • Created a data connection using the SQL connection string: Server=tcp:<dbname>.database.windows.net,1433;Initial Catalog=<dbname>;Connection Timeout=30;
   • Used username and password authentication type.
   • Tested the connection successfully using the "Test Connection" button.
3. Permissions:
   • Added the Azure ML Studio workspace as a Contributor over the Azure SQL Server resource.
4. Data asset import:
   • Followed the steps in the Microsoft documentation to import the data asset.
   • The data import job ran successfully, and the data asset appears in the Azure ML Studio "data assets" page.
5. AutoML job setup:
   • Selected the imported data asset during the AutoML setup.
   • Encountered the authentication error message mentioned above.

Thank you in advance for your help!

Answer 1

Hello Trevor,

Thanks for reaching out to us! I have seen this issue before, there is a troubleshooting guidance here - https://learn.microsoft.com/en-us/azure/machine-learning/how-to-troubleshoot-data-access?view=azureml-api-2#scriptexecutionstreamaccessauthentication

It can resolve most of the cases. Please take a look and have a try to see if this is helpful.

I hope this helps! Please let us know if you have any question.

Regards,

Yutong

-Please kindly accept the answer if you feel helpful to support the community, thanks a lot.