Fix adoption of orphan DataVolumes

[zcahana]

What this PR does / why we need it:

Fixes a bug in the DataVolumes adoption logic, which tries to patch a VMI (add an ownerReference) instead of patching the DV.

The observed behavior is that the DV is left unclaimed, which manifests itself in several ways:

• If the VMI is stopped, then it can't be started
• If the VMI is running, then it can't be stopped
• If the VM is deleted, then the deletion doesn't cascade to the DV.

Steps to reproduce (using this example VM):

# create any VM with dataVolumeTemplates
kubectl apply -f alpine.yaml

# delete the VM with with --cascade=orphan,
# i.e. release owned objects without deleting them
kubectl delete vm alpine --cascade=orphan

# recreate the VM (or another) referencing the same DV
kubectl apply -f alpine.yaml

# try to start the VM. It won't.
virtctl start alpine

Special notes for your reviewer:

Also added a unit test to cover this behavior.

Release note:

Fixes adoption of orphan DataVolumes

[ezrasilvera]

Thanks @zcahana . Can you please add to the description what is the observed error and how to produce it ?

[zcahana]

Can you please add to the description what is the observed error and how to produce it ?

Added.

[mhenriks]

/lgtm
/approve

[kubevirt-bot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: mhenriks

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 [mhenriks]

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

[zcahana]

@mhenriks Hi Michael, I've made a tiny change to the test as per @ezrasilvera's comment, can you please reapprove?

[ezrasilvera]

/lgtm

[zcahana]

@mhenriks do we need to backport this bugfix to previous releases? Based on the backporting policy it seems to be eligible for backporting, but is there any criteria to determine what eligible bugfixes are actually backported?

[davidvossel]

is there any criteria to determine what eligible bugfixes are actually backported?

There's no criteria. The policy is intentionally vague there. As it is today, anyone who wants to backport a bugfix can do that to any stable branch

[zcahana]

Very well. So based on the unformal criteria of "is anyone likely to hit this bug", I think we can skip backporting it.