devspace-plugin-loft VS microk8s

I'm biased but I do think what you're describing is a good use case for Kubernetes. I work for Loft Labs, we're the company that created vcluster. We do have a commercial product called Loft that lets you manage vclusters and offer them self-service to developers. If you want to get more info on that, the web site is loft.sh.

There are some products available, for example Loft who open sourced vcluster

We run on EKS and use https://loft.sh/ to deploy development environments. The engineer runs the single service locally that they want to modify. Any other services or databases that service connects to run remotely in our Development EKS cluster using Loft and port forwards to the engineers local environment.

Lastly, one day my employer will let my run Loft so that I can provide on-demand ephemeral k8s clusters to my dev + test teams.

Have a look at https://github.com/loft-sh/kiosk and maybe the paid version https://loft.sh/

How many friends? I would recommend evaluating Loft which is free for up to 3 users.

Automate the setup of new k8s environments, ideally they should be ephemeral, disposed regularly and rebuilt by the devs themselves, on demand. This will also keep costs in check, devs don't work 24/7 so why should their cloud infrastructure? A tool worth checking would be Loft

Loft solves this + much more. There are cost-saving features too, so it might actually pay for itself. Don't hesitate to book a demo.

Loft.sh provides self-serve Kubernetes solutions for cost optimization, CI/CD, policy enforcement, user management, collaboration, and more. It helps save on Kubernetes costs by using quotas and space constraints which helps while sharing your clusters among multiple users and teams. Auto delete for idle namespaces and sleep mode for idle workloads also saves costs.

FROM postgres:11CMD sleep 30 && postgres I'm reusing this example from the same issue with microk8s where I could solve it by changing the kubelet and kubeapi-server configuration files (see https://github.com/ubuntu/microk8s/issues/770 in case you're interested). I assume this is not possible with GKE clusters as they don't expose these files, probably for good reasons.

$ kubectl get podsNAME READY STATUS RESTARTS AGEbackend-deployment-66cff7d4c6-gwbzf 0/1 ImagePullBackOff 0 18s Before that it was ErrImagePull. So, my question is, how to tell it to use local docker images? Somewhere on the internet I read that I need to build images using microk8s.docker but it seems to be removed.

There's a whole community of hobbyists building Raspberry Pi clusters, porting things to work on various Arm processors, exploring and contributing to minimalist distros like k0s and microk8s, etc.

same pvc worked fine on "GKE" (Google Kubernetes Engine) but failing in my local cluster using microk8s

Local Kubernetes Clusters: It is now possible to run Kubernetes on local machines with tools such as Kubernetes in Docker (kind), minikube or MicroK8s. This allows developers to run their first experiments completely isolated from others and with low risk and low cost.

Local Kubernetes clusters are clusters that are running on the individual computer of the developer. There are many tools that provide such an environment, such as Minikube, microk8s, k3s, or kind. While they are not all the same, their use as a development environment is quite comparable.

Since the developer is the only one who has to access this cluster for development, local clusters can be a feasible solution for this purpose. Over time, several solutions have emerged that are particularly made for running Kubernetes in local environments. The most important ones are Kubernetes in Docker (kind), MicroK8s, minikube and k3s. For a comparison of these local Kubernetes options, you can look at this post.

Another very important distinction of this level is that developers have direct access to Kubernetes for the first time. While it is not strictly necessary, the standard case for the Kubernetes access with these tools is to use a local Kubernetes cluster, i.e. a Kubernetes cluster started with tools such as minikube, kind or MicroK8s on the local computer of the developer.