[Bug] mapt/pulumi takes too long to destroy the SNC instance

when destroying an SNC instance, one step takes very long to complete:

INFO @ destroying...............................................................................................................................................................................................................................................................................................................................................................................................................................................
INFO  -  aws:autoscaling:Group main-aos-asg deleted (430s)
INFO  -  aws:ec2:RouteTable routeTable-publicmain-aos-net0 deleting (0s)
INFO  -  aws:lb:LoadBalancer main-aos-lb deleting (0s)
INFO  -  aws:ec2:LaunchTemplate main-aos-lt deleting (0s)

I don't think this is normal.

This may be related to the OS shutdown, that should be skipped, but even if I destroy the instance with the flag checked from the AWS console, this destroy step takes ~300s.

Start by reproducing destruction of an SNC instance and inspect the Pulumi AWS resources involved, especially the autoscaling group and OS shutdown behavior. Done means the main-aos-asg destroy step no longer takes roughly 300–430 seconds, including when the AWS console’s shutdown option is enabled.