Add initial documentation for --vm-driver=none

[tstromberg]

Resolves #2575

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: tstromberg

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• OWNERS [tstromberg]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[bhack]

What it is the status of #2926?
Cause probably we can suggest a lighter solution for users that are looking for none.

[harpratap]

Please also add this - To make your locally built images run in minikube you need to tag your images with a version other than latest. Only then minikube will be able to pull those, the docker daemon is shared with minikube and your machine so no need to creating a separate repo.

[bw2]

When talking about running "within a customized VM environment" vs. "outside of a VM", what is a VM in this context? For example, is it bad to use --vm-driver=none on vanilla linux severs (with lets say CentOS7 or Ubuntu18)?

It would be helpful to specifically define "VM" and "developer machine".

[jakefeasel]

Sad as this documentation makes me, at least it does clarify the intent for this feature. Thanks.

On another note, it would be really nice if we could point people who are interested in running k8s directly on their servers to other projects which would help them accomplish that goal.

[tstromberg]

Sad as this documentation makes me, at least it does clarify the intent for this feature. Thanks.

On another note, it would be really nice if we could point people who are interested in running k8s directly on their servers to other projects which would help them accomplish that goal.

You've raised a good point about a valid use case. I've updated the docs in a way which hopefully captures the conditions in which you may want to run --vm-driver outside of a VM.

[bhack]

Probably there could be also some other alternatives to the insecure none like #3244

[RA489]

Also here, please provide reference to issue

[RA489]

@tstromberg please squash the commit to one.

[tstromberg]

@tstromberg please squash the commit to one.

Done, though I also don't feel it's explicitly necessary to squash commits.

[balopat]

LGTM, thank you