microk8s VS kubernetes-the-hard-way

And install microk8s:

One quick and easy win I can recommend, is microk8s.

microk8s currently has a showstopping issue that makes it guaranteed to have an irrecoverable failure in HA mode. see https://github.com/canonical/microk8s/issues/3227

k0s is better but also has a lot of bugs. it's the closest to vanilla kubernetes among all the distributions.

> like the simplest GPU support

linux users should be ready to install the nvidia device plugin. if they can't do that, they're never going to succeed in running a gpu accelerated application on their cluster anyway.

> like bootstrapping

in my experience, writing all the bootstrap scripts is painful. but now that there's chatgpt, so much of the drudgery as gone away.

MicroK8S to spin up a Kubernetes cluster

We will be using docker and microk8s from Canonical. For running our software during development, we will be using skaffold which is a great tool developed by Google.

I'm curious about this myself. Right now I'm considering putting microk8s on my Ubuntu server.

microk8s cluster, fully working Kubernetes in under 5 minutes

microk8s - https://microk8s.io

Another resource that I found pretty helpful was "Kubernetes the Hard Way" by Kelsey Hightower despite its complexity.

https://github.com/kelseyhightower/kubernetes-the-hard-way

I recently published Kubernetes the Harder Way, a guide loosely based on Kelsey Hightower's Kubernetes the Hard Way, but lenghtier, more explanatory, broader in scope, and - most importantly - harder, by targeting a local machine instead of Google Cloud Platform.

How about this? It will setup a kubernetes cluster on GCP with 3 worker nodes?

https://github.com/kelseyhightower/kubernetes-the-hard-way

Some costs here but click on the google cloud calc as it seems to have gone up since he wrote this:

If you wanna learn it deeply and quickly, Kelsey Hightower's Kubernetes the hard way is fantastic!

As for practicing, there’s k8s the hard way and similar flavours.

But if you boil down a control plane to its most essential components, it is basically a database (etcd), a webservice (apiserver) and two controllers (controler-manager, scheduler), none of which requires a Kubernetes cluster for their own needs. Self-hosting the control-plane the kubeadm way is absolutely not a requirement, you can also download the bare binaries and run them as basic systemd units out of the cluster. This is how the Kubernetes the hard way tutorial makes things work, if you want to have a look.