[BUG] crc start error message is not informative when leftover vm's

[maxandersen]

General information

Trying to get crc working again after not using it for some weeks.

download new CRC, on crc start I eventually get:

❯ crc start
INFO Checking if oc binary is cached
INFO Checking if podman remote binary is cached
INFO Checking if running as non-root
INFO Checking if HyperKit is installed
INFO Checking if crc-driver-hyperkit is installed
INFO Checking file permissions for /etc/resolver/testing
INFO Checking file permissions for /etc/hosts
ERRO Bundle 'crc_hyperkit_4.3.10.crcbundle' was requested, but the existing VM is using 'crc_hyperkit_4.3.8.crcbundle'

From a CRC user perspective that a VM is running is secondary info for me - might not even know a VM is behind it thus user is left with sense something is "broken" instead of "Hey, you seem to have reinstalled CRC and there are some leftovers - please cleanup to get it running".

Only because @praveenkumar helped me did I figure out that crc delete or crc cleanup could help me here

CRC version

crc version: 1.9.0+a68b5e0
OpenShift version: 4.3.10 (embedded in binary)

CRC status

ERRO The connection to the server api.crc.testing:6443 was refused - did you specify the right host or port?
 - exit status 1

CRC config

Host Operating System

ProductName:	  Mac OS X
ProductVersion: 10.15.3
BuildVersion:	  19D76

[zedr]

As a workaround, execute the following commands. Note that will delete everything regarding crc and Kubernetes on your system, including the VM, so make backups in case you need to retain some data.

crc delete
crc cleanup
rm -rf ~/.crc
rm -rf ~/.kube

This should fix the issue.

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[praveenkumar]

@gbraad we do need to discuss and plan for what should be the ux in that case.

[gbraad]

@code-ready/crc-devel Suggestion?

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[maxandersen]

Reopen?

[gorkem]

What is preventing us from deleting the old/invalid crcbundle during crc start instead of printing an error and failing?

[cfergeau]

Deleting the old/invalid bundle means removing the whole crc VM, ie data loss for the user. When there is a pre-existing VM using an old bundle, do we want to let the user run the older one with the newer crc, or do we want to force it to upgrade.
I think right now we are leaning toward the latter, but never quite finished it up, and the user experience is stuck somewhere in between the 2 options

[gorkem]

@cfergeau Is this waiting for selecting one of those options only?

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.