Docs: k8up removal

[SchoolGuy]

Summary

As "an administrator"\ I want "to be able to easily remove k8up"\ So that "I can reinstall it cleanly in case the operator messed up."

Context

During https://github.com/k8up-io/k8up/issues/834#issuecomment-1508012367 it became evident that some bugs are not triggered when deploying on a fresh cluster. As a clean re-deployment is IMHO not nice but an option I would like to know how to fully remove k8up so my FluxCD instance can reinstall it a few minutes later.

Out of Scope

No response

Further links

• https://github.com/k8up-io/k8up/issues/834#issuecomment-1508012367

Acceptance Criteria

• The page https://k8up.io/k8up/2.7/how-tos/uninstallation.html or a similar link is existing.
• The page lists how a full removal including the CRDs, Cluster Roles, ... is performed.

Implementation Ideas

No response

[SchoolGuy]

As context: I have a new problem with 4.2.2 that is unrelated to this issue but I am not sure how to fully reinstall k8up so that I have a fresh state. Tearing down the whole cluster is not an option for me.

[Kidswiss]

Hi @SchoolGuy

On top of my head:

• K8up creates a rolebinding in each namespace that a backup is active
• Everything else should be cleaned by uninstalling the helm chart and the CRDs

[SchoolGuy]

If it is that easy I will contribute docs in the evening.

[Kidswiss]

If I can find a bit of time I'll be looking into helm hooks that could help with automating that.

Especially for the update this seems to be a bit of an issue.