We recently upgraded from Azure ML Sdk V1 to Sdk V2, to get automatic support for HTTPS-Endpoints. Azure ML now doesn't use Container Instances anymore. By using Container Instances, we could shutdown our Endpoints to save costs over night. For that we were using the Azure Management API:
'https://management.azure.com/subscriptions/{subscriptionId}/resourceGroups/{resourceGroup}/providers/Microsoft.ContainerInstance/containerGroups/{containerInstance}/stop?api-version=2021-10-01'
Now using the sdk V2, there is no option to shutdown our Endpoint Resources. The only workaround to my knowledge is to delete this endpoint (Managed Instance) and recreate it the next day.
As a reference, the User Guide for Azure Machine Learning provides information on how to reduce costs, but it doesn't cover Endpoints.
Since the deployment takes way longer now, I was wondering, if there is another option to reduce our costs without deleting endpoints?
Let me know, if you need more information on my issue. Thanks in advance for your help :)
