As someone who's been using, administering, and consulting on Kubernetes since 2017, this list is bang on, with maybe one major omission.

I could not imagine my day to day workflows without tools like kubectx/krew/k9s.

What I think the author doesn't really touch upon is how much of a revolution GitOps has been for Kubernetes. I've watched it grow from idea to stable implementation and in 99.9% of cases if you are using Kubernetes but not doing GitOps (preferably via FluxCD or ArgoCD) I know exactly the sort of mess I'm going to find in your clusters and organisation.

Kyverno is also another tool that has quickly become indispensable. What's not so obvious about it is you can lean on it for all the things you'd lean on bash in a typical non-k8s/non-cloud-native infra. It can be used as the de facto Kubernetes duct tape, not just for policies.

Kubernetes without a proper Gitops (not half Gitops half manual) will devolve into a mess, similar with cloud infra without IaC, this needs to be front and center.

I'll take this moment to remind everyone that you don't have to use k8s, if you don't need it.

Regular VMs are still a great and simple solution to many, many, many deployment scenarios.

The same can be said for pretty much any shiny tool (FaaS, Rust, React, GraphQL, etc). These tools are sometimes the right solution, but they sometimes aren't.

Then you realize you need monitoring, then logs, then containers, then internal DNS, then service discovery, then H/A storage, then... I've gone down this road with virtual machines then docker swarm and to an extent early nomad.

Slightly off topic, but the author should really reconsider the flashing title and logo. It is incredibly distracting.

As a counterpoint I kinda liked it. It doesn't continue down the page but I thought it was a nice touch.

anybody else enjoy direnv for managing multiple eks clusters? just cd into a said directory and it sources needed kubeconfig - no need for fancy mergeing setups, ect

Good read