Support GVK reconciliation for a certain time after RGD is deleted

[michaelhtm]

Feature Description

Problem Statement:
When RGDs are deleted, their instances/resources are abandoned orphaned, unless the allow-crd-deletion flag is enabled.

Proposed Solution:
I'd like kro to be able to manage my instances for a certain amount of time, even after the RGD is deleted (some sort of grace period) before stopping reconciliation of the GVK

• Please vote on this issue by adding a 👍 reaction to the original issue
• If you are interested in working on this feature, please leave a comment

[bschaatsbergen]

One idea that could be worth exploring is using Kubernetes finalizers to block deletion until certain conditions are met. Projects like ArgoCD use this pattern, which might provide the kind of grace period we’re looking for. I don’t think KRO currently supports finalizers on RGDs, but it could be an interesting approach to consider.

[jakobmoellerdev]

Hey folks, would love feedback and ideas on #763 to get this solved

[tjamet]

I do think we should use finalizers and match k8s standard behaviours by default (i.e. delete unless no-cascade)
But this is, as @jakobmoellerdev mentions, a discussion to have in #763