microsoft / MLOpsPython

MLOps using Azure ML Services and Azure DevOps
MIT License
1.2k stars 1.1k forks source link

Keeping AZURE_RM_SVC_CONNECTION and WORKSPACE_SVC_CONNECTION same #403

Open jyotiranjansharma11 opened 2 years ago

jyotiranjansharma11 commented 2 years ago

Bumping up #195

My understanding is, if I have a Service Connection of Subscription level, (as it is a superset of the resources inside it) it should also be able to do modifications over resources like Azure ML Workspace.

Basically I am keeping the AZURE_RM_SVC_CONNECTION and WORKSPACE_SVC_CONNECTION same in the variable group.

The job: "Run_ML_Pipeline" results in an invalid http request:

URL https://[missing region].aether.ms/api/v1.0/subscriptions/xxx/resourceGroups/[missing rg]/providers/Microsoft.MachineLearningServices/workspaces/[missing workspace]/PipelineRuns/PipelineSubmit/xxx for HttpRequest is not correct.

So can I not keep AZURE_RM_SVC_CONNECTION and WORKSPACE_SVC_CONNECTION same in the variable group?