Graceful shutdown of Kubernetes cluster

what about backing up the etcd config to restore on start up?

https://coreos.com/etcd/docs/latest/v2/admin_guide.html

Doesn't that save you quite some time?