KevsRobots Learning Platform
35% Percent Complete
By Kevin McAleer, 2 Minutes
Page last updated May 24, 2025

Even though K3s simplifies Kubernetes installation, things can still go wrong β especially with Raspberry Pis.
This lesson walks you through common issues and how to troubleshoot them effectively.
Worker node installs, but doesnβt show up with kubectl get nodes.
Ensure youβre using the correct master IP:
ping <MASTER-IP>
Check the K3s agent logs on the worker:
journalctl -u k3s-agent -e
Worker node fails to join with error about invalid token.
Re-check token syntax:
sudo cat /var/lib/rancher/k3s/server/node-token
If needed, regenerate token by reinstalling K3s on master with:
/usr/local/bin/k3s-uninstall.sh
K3s or agent fails to start on boot.
Check logs:
journalctl -u k3s -e
journalctl -u k3s-agent -e
Common fixes:
Ensure cgroup support is enabled:
Check /boot/cmdline.txt or /boot/firmware/cmdline.txt has:
cgroup_memory=1 cgroup_enable=memory cgroup_enable=cpuset
Check for disk space:
df -h
Restart the service:
sudo systemctl restart k3s
Worker tries to join but fails due to DNS or unreachable host.
/etc/hosts on each node.On master:
sudo /usr/local/bin/k3s-uninstall.sh
On worker:
sudo /usr/local/bin/k3s-agent-uninstall.sh
Then reinstall as per earlier lessons.
| Step | Command |
|---|---|
| Check node status | kubectl get nodes |
| View K3s logs | journalctl -u k3s -e |
| View agent logs (worker) | journalctl -u k3s-agent -e |
| Check config | cat /etc/rancher/k3s/k3s.yaml |
| Inspect cluster components | kubectl get pods -A |
Once your cluster is running and stable, itβs time to start using it!
Next up: Using Your K3s Cluster
You can use the arrows β β on your keyboard to navigate between lessons.
Comments