This is why I'm bearish on Kubernetes for most organizations. Not because it doesn't work great, but because its complexity is so high. I have this adage about

This is why I'm bearish on Kubernetes for most organizations. Not because it doesn't work great, but because its complexity is so high. I have this adage about not solving problems you don't have, but it's the same concept in the article. Kubernetes is for Google. And I guess Netflix could use kube. Amazon. Facebook, sure. But the complexity overhead is so enormous that you are probably causing more problems than you're solving, especially if you're already using something that works (since this is the same exact thing as rewriting the software from scratch, but for sysops). Alas. New and shiny is a powerful temptation.