1

I have this "issue", that I don't really know if it is the supposed behaviour or not. I'm running some services in Google AppEnginge cloud platform. I apply this app.yaml config for automatic scaling: instance_class: F1 automatic_scaling: min_instances: 0 max_instances: 2 min_idle_instances: 0 max_concurrent_requests: 50 This service is intended for development and testing so I don't mind that it does scale down to zero. But thing is, as soon as the service is deployed, no instance is running. After making the first request, instance is fired up and once it's running, even though no new traffic is being served it does not get destroyed. If billing is not applied to it, I don't care, but I cannot realize if it's being billed or not while it's note recieving traffic. Documentation is kind of ambiguous with this. Any tips or clues?

Luciano César Natale
  • 65
  • 2
  • 10
  • Heya @Luciano César Natale! If the answer is good for you, could you accept it to give it more visibility over the community? Other users with the same doubt may benefit from it! – Pablo Almécija Rodríguez Sep 26 '19 at 14:51

1 Answers1

1

For automatic_scaling, if you declare min_idle_instances: 0, all your instances are dynamic instances, that is, instances will be started and shut down depending on their needs (e.g. more or less traffic). In terms of billing, this means that you will stopped being charged fifteen minutes after the last request has finished processing, as documented here. In any case, you will only be charged for the min_idle_instances specified. As you specified 0, you will not be charged for it.

To verify the kind of availability of your instances, you can go to you App Engine service in the Cloud Console, click on the desired version and then in the number of instances.

As for why you see no instance is running despite the fact that your application is already serving, the Cloud Console takes some time to refresh that information.

Pablo Almécija Rodríguez
  • 176
  • 3