Add Worker Nodes to K3s Cluster

Add Workers

Use this tutorial after the control plane is operational to scale out your cluster with additional worker nodes.

Step 1: Add the node to inventory

Update ansible/inventory/hosts.yaml with the new worker node details under k3s_agents:

k3s_agents:
  hosts:
    worker-01:
      ansible_host: WORKER_TAILSCALE_IP
      ansible_user: sudhanva
      k3s_node_ip: "WORKER_TAILSCALE_IP"

Replace WORKER_TAILSCALE_IP with the output of tailscale ip -4 on the worker. Every node uses its Tailscale IP as the node IP, so nodes reach each other and the API server over the tailnet regardless of LAN addressing.

Step 2: Run the agent provisioning role

Execute Ansible against the target worker:

ansible-playbook -i ansible/inventory/hosts.yaml ansible/site.yaml --limit worker-01

The playbook installs prerequisites, joins the K3s cluster using the cluster join token, and starts the k3s-agent service.

Step 3: Verify the new node

Confirm that the new worker reports Ready and displays its role:

kubectl get nodes -o wide

Check that the Flannel CNI and workload pods are scheduling onto the worker:

kubectl get pods -A -o wide