After dealing with different enterprise systems in AWS for a few years now, I'd say that this is what AWS ELB feature to remove a system out of an ASG as part o

After dealing with different enterprise systems in AWS for a few years now, I'd say that this is what AWS ELB feature to remove a system out of an ASG as part of its lifecycle exists for. If you aren't using similar horizontal scaling strategies, you have a disaster waiting to happen basically and basically deserve an outage you can't diagnose / RCA well. If you can't ssh in afterwards that means your provisioning for automation / instrumentation and you should stay be able to isolate attention to that part of an instance's lifecycle.