feat: update RGD controller to set CRD ownership and optimize watches

[jakobmoellerdev]

Updated the ResourceGraphDefinition controller to set correct ownership references on CRDs and switched to metadata-only watches for improved performance and reduced memory consumption. Added integration tests to validate CRD ownership.

This would resolve both cleanup of the CRD with background deletion mode and resolution of owner refs in general but has the side effect that RGDs must stay non-namespaced (I think this is static by now and we can move to ownerrefs safely)

resolves #745

[k8s-ci-robot]

Skipping CI for Draft Pull Request.
If you want CI signal for your change, please convert it to an actual PR.
You can still manually trigger a test run with /test all

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by: jakobmoellerdev
Once this PR has been reviewed and has the lgtm label, please assign barney-s for approval. For more information see the Code Review Process.

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

Needs approval from an approver in each of these files:

• OWNERS

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

[a-hilaly]

i'm not very sure we need ownerReferences on CRDs. This is conflicting with the fact that we have a --allow-crd-deletion flag https://github.com/kubernetes-sigs/kro/blob/main/helm/values.yaml#L97-L98.

[jakobmoellerdev]

I think we have very static / classic controller ownership relations between RGD<->CRD. Thus I think also deletion behavior can be controlled by an ownerref and allow crd deletion can be deprecated in favor of ownerreferences. just my opinion of course, happy to hear other thoughts.

[dudo]

Thanks for picking up my issue, @jakobmoellerdev! I'm obviously biased, but the owner relationship feels logical to me. Why would we want to keep CRDs around after an RGD is deleted?

[tjamet]

I would say we do want to reconcile RGDs when the CRD they manage gets deleted so we re-create them