Repository navigation
Deleting CRD results in could not list instances: unable to find a custom resource client for addresspools.metallb.io: unable to load root certificates: unable to parse bytes as PEM block #1501
Description
I'm getting the following error when trying to delete the addresspools or bgppeers CRDs.
% kubectl describe customresourcedefinitions.apiextensions.k8s.io addresspools.metallb.io
...
Client Config:
Ca Bundle: Cg==
...
Conditions:
Last Transition Time: 2022-07-14T08:19:58Z
Message: no conflicts found
Reason: NoConflicts
Status: True
Type: NamesAccepted
Last Transition Time: 2022-07-14T08:19:58Z
Message: the initial names have been accepted
Reason: InitialNamesAccepted
Status: True
Type: Established
Last Transition Time: 2022-07-14T08:20:36Z
Message: could not list instances: unable to find a custom resource client for addresspools.metallb.io: unable to load root certificates: unable to parse bytes as PEM block
Reason: InstanceDeletionFailed
Status: True
Type: Terminating
Installed MetalLB using the metallb Helm chart v0.13.3 from https://metallb.github.io/metallb.
I then deleted the helm install shortly after when I realized it was installed to the wrong namespace.
Running k3s: v1.23.8+k3s2
Using Cilium v1.11.6
Presumably, this is because I deleted the controller before the CA could be populated here. Assuming that's the case, it's probably worth putting a "real" dummy CA so the CRD can be deleted if the controller never started.
Seeking some advice on whether it's safe to force delete the CRD and to see if adding a CA is a valid solution.
Activity
fedepaol commentedon Jul 14, 2022
So just to understand better the steps, you just installed and deleted the chart right?
optix2000 commentedon Jul 14, 2022
In rapid succession, before any of the pods had started.
fedepaol commentedon Jul 14, 2022
Understood, thanks for the report
optix2000 commentedon Jul 14, 2022
Literally