Comment # 14 on bug 1171770 from Richard Brown

(In reply to Francisco Freitas from comment #13)
> The error is also present on multi-master cluster. I deployed a cluster from
> the release 20200516 using the following commands : 
> 
> ```
> kubicctl init --haproxy loadbalancer --multi-master
> loadbalancer.cluster.local
> kubicctl node add --type master master02
> kubicctl node add --type master master03
> kubicctl node add worker01
> ```

So..
I've used kubeadm init --image-repository to use only upstream containers -
problem still occurs
I've used rebuilt kubernetes 1.18.2 containers - problem still occurs
i've deployed it on kubernetes 1.17.5 - problem still occurs
I've used only upstream weave, cilium and other CNI providers - problem still
occurs
I've used https://download.opensuse.org/history/ to move my nodes to every
version of kubic we've had in May - problem still occurs

I'm officially flummoxed - does anyone have any idea when this last worked for
sure? because I'm running out of things to rule out