Unable to deploy to machine learning workspace online endpoint

Ben 0 Reputation points
2025-02-18T20:27:33.6533333+00:00

I am unable to deploy/update my online endpoints through in my machine learning workspace. The workspace has its VNET managed by Azure and I'm getting errors related to the subnet. I've ensured that the workspace has the Network Contributor role on the resource group.

2025-02-18T04:20:14Z Received managed network subnet creation error for Workspace VNET, please check workspace network settings and retry. Error: ErrorDetail: Code:UserError, Message:A validation error has occured.. Managed network subnet creation details: ManagedNetworkCreationCreationStartTime (UTC): 02/18/2025 04:16:36 Subscription ID: ***** ResourceGroup: ***** Workspace: ***** NetworkId ID: dd2cb742-5fbb-4f91-be76-07a30f087319 SubnetName: 0cdc02dd3a0f434c9e7977026e820398 NetworkBoundaryName:  NetworkComputeType: AMLCompute\",\"Data\":{\"ErrorType\":\"ManagedNetworkError\",\"IsTerminal\":true},\"InnerException\":null,\"HelpURL\":null,\"StackTraceString\":\"   at Microsoft.MachineLearning.InferenceDeployment.Silo.Grains.Flows.CreateManagedNetworkSubnetFlow.<>c__DisplayClass7_0.<<ExecuteTurnAsync>b__2>d.MoveNext() in /mnt/vss/_work/1/s/src/Silo/Grains/Flows/CreateManagedNetworkSubnetFlow.cs:line 233\\n--- End of stack trace from previous location ---
Azure Machine Learning
Azure Machine Learning
An Azure machine learning service for building and deploying models.
3,141 questions
{count} votes

1 answer

Sort by: Most helpful
  1. Amira Bedhiafi 28,851 Reputation points
    2025-02-18T22:03:56.2166667+00:00

    Your Azure Machine Learning workspace doesn't have the correct subnet configuration when it comes to sufficient IP addresses and proper delegation to Microsoft.MachineLearningServices/workspaces. You need to verify if the workspace has the "Network Contributor" role on the resource group (that's what you did) but indeed don't forget to check for any NSGs or firewalls blocking required traffic.

    Retry the deployment after addressing these issues, and if the problem persists, you can contact Azure Support for further assistance.

    0 comments No comments

Your answer

Answers can be marked as Accepted Answers by the question author, which helps users to know the answer solved the author's problem.