Microsoft Azure Machine Learning support team blog https://docs.microsoft.com/archive/blogs/azuremlsupport/feed.xml Wed, 17 Apr 2019 03:24:14 GMT en-US hourly 1 ML workspace creation fails with error: "The location constraint is not valid" https://docs.microsoft.com/archive/blogs/azuremlsupport/ml-workspace-creation-fails-with-error-the-location-constraint-is-not-valid Tue, 12 Jan 2016 01:12:03 GMT https://blogs.technet.microsoft.com/azuremlsupport/2016/01/12/ml-workspace-creation-fails-with-error-the-location-constraint-is-not-valid/

As of today the Azure machine learning service is offered in three locations around the world:

South Central US,

West Europe,

Southeast Asia

When you create a ML workspace, it will need access to an Azure storage account in the same location. If you do not supply an existing storage account, then it will try to create one on the fly.

In case your Azure subscription does not have access to the chosen ML workspace location, then the storage account (and subsequently the ML workspace) creation will fail with the error: "The location constraint is not valid".

You can test which locations your Azure subscription has access to by looking at the create new storage account template. It will list all locations that your subscription has currently access to:

Note that in above screenshot I have access to 7 locations that include all three AML locations: South Central US, West Europe, Southeast Asia. Thus I can create ML workspaces in all three locations and won’t see the issue.

The same may not be true for your subscription. Many Azure subscriptions may by default just have access to 4 locations. And these may not include your desired ML workspace location.

If you find that you cannot create an ML workspace in your chosen location because your Azure subscription has no access to it, then you will need to open a support ticket and ask for the desired locations being added to the Azure subscription. When opening the ticket please make sure that you guide it towards Azure subscription management.

Do you just want to do quick testing of the Azure machine learning service? Then we have another solution for you. Under the URL: https://studio.azureml.net/  you can create a free workspace that does not require any Azure subscription. 

]]> “User is not authorized to access this workspace” error and Workaround https://docs.microsoft.com/archive/blogs/azuremlsupport/user-is-not-authorized-to-access-this-workspace-error-and-workaround Tue, 08 Sep 2015 15:24:47 GMT https://blogs.technet.microsoft.com/azuremlsupport/2015/09/08/user-is-not-authorized-to-access-this-workspace-error-and-workaround/ Azure Machine Learning (ML) Service has recently enabled the support of Azure Active Directory accounts. This change may impact the Azure ML Workspace created prior to the change, and the Workspace has one or more owner or user account using a corporate email address. Those users may lose access to the Workspace. The following error message will be returned when the user tries to connect to the Workspace from Azure Management Portal, even though the Workspace is listed there under Machine Learning with status of online.

User:xxxxx@corporatedomain.com

Workspace ID:xxxxxxxxxxxxxxxxxxxxxxxxxxxxx

The Signed-on user is not authorized to access this workspace. Please try to sign out and re-sign on with an authorized user account.

If the user goes to https://studio.azureml.net directly, and sign in using his or her corporate email address, the Workspace is no longer in the Workspace list on the Studio site.

To regain access to the Workspace for these impacted users, please take the following steps

  1. Go to Azure Management Portal, change the workspace owner from the corporate email address to a valid live.com or Hotmail.com account
  2. Launch InPrivate browser, go to https://studio.azureml.net
  3. Log in with the live.com or Hotmail.com account
  4. Invite the corporate email address from the Azure ML Workspace as User or Owner as appropriate

]]>