So in recent versions of Kubernetes, when pods are killed from a drain event, they go into a state of Failed with the reason being Terminated. They hang out there for a time, but will eventually get GC’d. This can however mess with some stats and it looks kinda scary if you aren’t prepa…