I was following this guide for setting up Kubernetes on AWS. At the end of the guide, I was running 1 master and 3 minions using AWS EC2 machines. Then I decided to shut them all off - thinking if the master is down, none will revive.
After few hours, I find 4 minions running.
Could it be the case that a master spawn up as a minion - somehow the order of operations caused this to happen?
How do I safely kill them all?
To safely shut down your cluster you should run cluster/kube-down.sh as described at the end of the guide you linked to.