0

I have created a deployment on ACI with Azure ML service, and its status is healthy.
When deploying, I set auth_enabled=True, so that the service requires authorization keys to respond.

I can get the service auth keys for that deployment in my Azure ML service workspace ws in a Python console via

from azureml.core.webservice import Webservice
services = Webservice.list(ws)
services[0].get_keys()

However, it would be convenient to access to this information through Azure Portal or the Cloud Shell.

In Azure Portal (differently to what happens for AKS) there's no auth fields shown, also when accessing Advanced Settings by trying to edit the deployment:

enter image description here

Can you suggest ways to access those credentials?

Davide Fiocco
  • 5,350
  • 5
  • 35
  • 72

1 Answers1

1

You can add the Azure Machine Learning CLI to the cloudshell by running

az extension add -n azure-cli-ml

after that, you can retrieve the keys for any service by running

az ml service get-keys --name magical-service --workspace-name davide-workspace
csteegz
  • 380
  • 1
  • 5