Welcome to the Linux Foundation Forum!

Lab 4.1 Issue upgrading worker to version 1.22

ojmendezz
ojmendezz Posts: 4
in LFS258 Class Forum

I've got this error when upgrading the version of kubeadm in the worker node to 1.22.1-00:

Unable to fetch the kubeadm-config ConfigMap: failed to decode cluster configuration data: no kind "ClusterConfiguration" is registered for version "kubeadm.k8s.io/v1beta3" in scheme "k8s.io/kubernetes/cmd/kubeadm/app/apis/kubeadm/scheme/scheme.go:31"

After searching on Internet I found this was reported as an issue and was fixed. So, just upgraded the whole cluster to the latest patch version: 1.22.6-00 and worked well.

I guess the lab instructions will be updated soon to use the 1.23 version anyway.

Comments

  • chrispokorni
    chrispokorni Posts: 2,606

    Hi @ojmendezz,

    I have seen this behavior once before, and it was related to an incompatibility between kubeadm.k8s.io/v1beta2 from the early version and kubeadm.k8s.io/v1beta3 referenced by the updated version.

    Regards,
    -Chris

  • fabio.sasso
    fabio.sasso Posts: 3

    @chrispokorni said:
    Hi @ojmendezz,

    I have seen this behavior once before, and it was related to an incompatibility between kubeadm.k8s.io/v1beta2 from the early version and kubeadm.k8s.io/v1beta3 referenced by the updated version.

    Regards,
    -Chris

    I confirm this, I got the same error trying to join a cluster create with kubeadm 1.22 with a worker running kubeadm 1.21

    THANKS!

  • chrispokorni
    chrispokorni Posts: 2,606

    Hi @fabio.sasso,

    The recommendation is to join a cluster with components on the same version across nodes.

    Regards,
    -Chris

Categories

Upcoming Training