[AVM Module Issue]: Machine Learning Module should add the creation of Microsoft.MachineLearningServices/workspaces/connections

[Menghua1]

Check for previous/existing GitHub issues

• I have checked for previous/existing GitHub issues

Issue Type?

Feature Request

Module Name

avm/res/machine-learning-services/workspace

(Optional) Module Version

No response

Description

Please add the creation of Microsoft.MachineLearningServices/workspaces/connections.

For example (azd):

• https://github.com/Azure/azure-dev/blob/main/templates/common/infra/bicep/core/ai/hub.bicep#L69

CC: @jongio

(Optional) Correlation Id

No response

[avm-team-linter]

@Menghua1, thanks for submitting this issue for the avm/res/machine-learning-services/workspace module!

Important

Please note, that this module is currently orphaned. The @Azure/avm-core-team-technical-bicep, will attempt to find an owner for it. In the meantime, the core team may assist with this issue. Thank you for your patience!

[AlexanderSehr]

Hey @Menghua1, thanks for the issue.
The request makes absolute sense, hence I removed the 'Needs: Triage' label. However, as the module currently has no owner that would automatically be assigned to this issue, it may take a moment for the community to work on it. If you're up for the task you can of course also create the contribution yourself. 💪

[matebarabas]

Good news, this module just got a new owner! @cecheta, can you please look into this issue? Thanks!

CC: @ross-p-smith as secondary owner

[cecheta]

Thanks for raising this, we'll look into this in due course

[cecheta]

Hi @Menghua1 , it looks like Microsoft.MachineLearningServices/workspaces/endpoint is deprecated now? I receive this error:

{
  "status": "Failed",
  "error": {
    "code": "DeploymentFailed",
    "target": "/subscriptions/***/resourceGroups/rg-dev/providers/Microsoft.Resources/deployments/main",
    "message": "At least one resource deployment operation failed. Please list deployment operations for details. Please see https://aka.ms/arm-deployment-operations for usage details.",
    "details": [
      {
        "code": "ValidationError",
        "message": "Endpoint resources are no longer supported, and have been migrated to connections. Please delete this resource and re-create it as a connection.",
        "details": []
      }
    ]
  }
}

Would it be ok to update the issue to just include connections?

[Menghua1]

@jongio Any ideas about above comments? Could we consider only including connections in AVM?

[jongio]

@jongio Any ideas about above comments? Could we consider only including connections in AVM?

Please track down where this module is being use, who added it to infra/core, and then figure out a path forward.

[Menghua1]

@jongio Any ideas about above comments? Could we consider only including connections in AVM?

Please track down where this module is being use, who added it to infra/core, and then figure out a path forward.

We did not find repo to use this module and @wbreza added this module to infra/core.

During the test, it looks like Microsoft.MachineLearningServices/workspaces/endpoint is deprecated now. Do you have any ideas about this?@wbreza

[jtracey93]

@jongio, from the comments/chat here from @Menghua1 it looks like this one is maybe not needed, can you and the team confirm?

#RR

[jongio]

Let's see what @wbreza says.

[wbreza]

endpoints are no longer supported and all new deployments must be migrated to connections. An example of this after a migration can be found in the AZD AI Studeio Starter template

[jongio]

@Menghua1 we can close this.

[Menghua1]

@Menghua1 we can close this.

This issue cannot be closed yet, because not only is the endpoints missing, but also a connections is missing. I removed the unsupported endpoints and updated the description.

[Menghua1]

@cecheta I've updated the issue to include only the connections.

[jtracey93]

@cecheta are you able to provide an ETA on resolution for adding this to the module for @jongio, please?

Thanks

Jack

[cecheta]

I will start looking at it now, however I am also working on the new avm/ptn/ai-platform/baseline module, and I am on leave for ~2 weeks from the start of August.

Hopefully this will be implemented before the end of August.