I've worked with Kubernetes extensively since it became popular. Of the maybe hundred of companies I've worked with trying to onboard to Kubernetes, I don't th

I've worked with Kubernetes extensively since it became popular. Of the maybe hundred of companies I've worked with trying to onboard to Kubernetes, I don't think a single one really had a use case that justified the increased complexity. Kubernetes seems so great until you're suddenly drowning in YAML, un-upgraded clusters out of their support window, the need for an unbelievable amount of extra tooling (e.g. ArgoCD, Helm, CSI plugins, Operators), etc. Plenty of places made it work by paying for huge teams to manage it, but deploying software on regular old Linux servers is fine for most people with just a small amount of discipline. Use boring technology.