Beginning again with the great work left off on from the previous post – I have started up fixing errors.

Two things here – kubeadm is trying to reach out to the Internet. Not sure if that is going to cause any complications later. For the time being I am going to to disregard that error.

# kubectl version 
Client Version: v1.32.0
Kustomize Version: v5.5.0

# kubeadm ini --kubernetes-version v1.32.0

Appears the kubeadm wants to generate new manifests. I simply moved the ones located in /etc/kubernetes/manifests to a backup folder just in case.

THIS IS THE MANUAL WAY
# mkdir /etc/kubernetes/manifests/backup

# mv -t /etc/kubernetes/manifests/backup etcd.yaml kube-apiserver.yaml kube-controller-manager.yaml kube-scheduler.yaml

PREFERRED WAY
# kubeadm reset

Seems to be a healthy initiation. Until… it wasn’t.

the kubelet-check hit an error – attempting to allow port access on 10248 to see if that fixes the issue.

# ufw allow 10248

Still nada!

Removed the config.toml file from /etc/containerd and then ran

# systemctl restart containerd && systemctl restart kubelet

# systemctl status containerd  && systemctl status kubelet

Both appear to be running fine now.

DISTRACTION!

Just out of curiosity I am going to see how this works on a standard Ubuntu Server install – I feel like there is either missing dependencies or a change with Kubernetes v1.32 and container runtimes that I am not tracking yet. I’ve read a few things about a transition away from the pre-installed docker engine. Plus a break would be nice!

Since I already had a working Ubuntu Server on node1 I went ahead and cloned it to node3. Later I will attempt to make another template for Ubuntu installs – something was off on the last one. I was surprised how easily Proxmox cloned and sent the OS to the other node. Just need to change the hostname now and should be good to go.

# hostnamectl

Looks like the Machine IDs are identical curious how that will conflict with anything after changing the hostname. It can be edited here: /etc/machine-id.

UPDATE: The Machine ID determines DHCPs IP address issued to the VM. I changed the first two numbers from 72 to 69, then reboot. DHCP then issued a unique IP address to the VM.

# hostnamectl hostname lama2

Then change from old hostname to new hostname by editing the hosts file with nano:

I don’t believe cloud-init was applicable to this install. I believe I went through manual installer a while back. I am going to change it anyway on recommendation of information I was reading on Linuxize.

# sudo nano /etc/cloud/cloud.cfg

Making a backup before doing any major changes to my VMs is my new favorite thing! It’s simply a lifesaver!

Proxmox backup of the VM to my Synology server. I need more storage space. I am going to get really backup happy.

After a brief delay, I reconfigured the Machine-ID for the VM to be issued a new IP address from DHCP. Perhaps I need to get that template setup sooner. 😀

Time for sleep! Until tomorrow!