I'm not sure if I can agree with the author. We moved from plain AWS EC2 to K8 to improve our continuous delivery pipeline and it worked. Before we had dozen of

I'm not sure if I can agree with the author. We moved from plain AWS EC2 to K8 to improve our continuous delivery pipeline and it worked. Before we had dozen of custom terraform scripts, custom runners on gitlab, and it became much cleaner now with K8. Other aspect that improved was monitoring and introspection. I think mainly because K8 offers a reach set of tools and patterns helping exactly in that matters. It comes with its own costs, but I can definitely say that we went from a tailor made solution to a standard set of techniques and tools, which is a huge improvement, IMO.