I was wondering about the minimum number of instances I can select when using auto-scaling with Azure App Service. I have previously used Google Cloud Platform, and became quite familiar with both Google Kubernetes Engine and Google App Engine. The latter allows you to scale down to 0 instances when the app/service is not in use, which is great for apps/services with intermittent loads. It was also very fast to go from 0 to 1 instance, when a request is made to the app/service.
I have poked around the Azure docs, but I haven't found a way to do something similar with Azure as of yet. Can I have an Azure App Service - or other managed service - scale down to 0 instances when there is no traffic (i.e. all of our clients are located in North America, and so there will be little to no interaction with our app at night, meaning it would be ideal for the apps/services to go dormant at night unless a request comes in to their URL where a small start-up time for the container/instance would be acceptable).