vm-driver-none fallback detection

[chris-hammond]

It looks like minkube changed the error code returned when trying to run minikube docker-env using --vm-driver='none' from 64 to 51. (Why??!!) For obvious reasons, this breaks skaffold when used with minikube in this configuration.

Expected behavior

When using skaffold deploy with older versions of minikube (e.g. v1.12.3) running --vm-driver='none', the following message is displayed:

WARN[0000] Could not get minikube docker env, falling back to local docker daemon: getting minikube env: running [/usr/local/bin/minikube docker-env --shell none -p minikube]
 - stdout: ""
 - stderr: "* 'none' driver does not support 'minikube docker-env' command\n"
 - cause: exit status 64
Fallback should function the same with the latest version of minikube.

Actual behavior

The following error occurs when running skaffold deploy:

creating runner: creating builder: getting docker client: getting minikube env: running [/usr/local/bin/minikube docker-env --shell none -p minikube]
- stdout: "\n\n"
- stderr: "X Exiting due to ENV_DRIVER_CONFLICT: 'none' driver does not support 'minikube docker-env' command\n"
- cause: exit status 51

Information

• Skaffold version: 1.14.0 and 1.15.0
• Minikube version: 1.13.1

Steps to reproduce the behavior

Using any skaffold project with minikube v1.13.1 running as --vm-driver='none', run skaffold deploy.

[briandealwis]

This was changed with kubernetes/minikube#9126 in minikube 1.13.0.