Lab 2.2. Deploy a New Cluster - cant get worker node to join the cluster

alizc

Hi Im using ubuntu 20.04 on my two VMs (one cp, one worker node).
Im trying to execute the sudo kubeadm join command on the worker node

The CP node seems to work fine as I can see both the cp and worker node when executing kubectl get node (though worker node's status is NotReady). The cp node is also active when I run sudo systemctl status kubelet.

I have tried to run sudo kubeadm reset on my worker node and then the join command again but i get the same output.

alizc

I have made some progress where i got the woker node to join the cluster
however i have not managed to get it's status to be "ready"

could this be due to removing /etc/cni/net.d ? it said to do so on the terminal when doing kubeadm reset on the woker node?

chrispokorni

Hi @alizc,

A successful join operation depends on several key factors:

• Type of VMs - which cloud provider is used or what local hypervisor provisioned the virtual machines?
• Virtual Resources Profile - the size of the VMs in terms of CPU, RAM, disk size, and the types of network interfaces (nat, host, bridged).
• Private IP Address Range of the VMs' subnet - what private IP addresses are assigned to the VMs.
• Operating System - Ubuntu 20 is quite outdated, and the lab guide recommends using a more recent Ubuntu LTS release.
• Firewalls - Improperly configured cloud VPC firewalls/security groups or hypervisor's promiscuous mode can impact connectivity.

Additionally, it’s important to note that the lab guide no longer uses Kubernetes version 1.31.1, but a more recent Kubernetes release instead.

Please provide the details noted above to help with troubleshooting your cluster.

Regards,
-Chris

alizc

Hi thanks for the response. I solved the problem by creating a new vm and just ran the scrips with the same VM setup as before. Not too sure what the actual issue was