Comment by ghshephard

Okay - I'm intrigued. How do you recommend managing infrastructure which consists of thousands of servers? With k8s - you just delete the pod and it gets recreated with a clean image within a few seconds. What's your proposal?