Hello, we have encountered upgrading issue with upgrading rke2 to 1.31.x due to this: https://github.com/kubernetes/kubernetes/issues/129385 https://github.com/kubernetes/kubernetes/pull/124220#discussion_r1896525461 apparently k8s is not supporting in-place upgrades which is what we have been doing with rancher managed rke2 clusters for a quite a long time... I know that draining configuration is also possible which should mitigate this issue but this is a big shift for us as this makes upgrade very slow for large clusters and also potentially disruptive as we have some workloads that can only be disrupted per AZ. So I guess my question is what is the rke2 stance on in-place upgrades?

We've never enforced that via any technical measures, just like we don't enforce any restrictions around not skipping minors... however if it is not supported by upstream, we don't support it either.