[k8s]Troubleshooting - worker node failure

1. Fix the broken cluster

k get nodes

systemctl status docker

systemctl status kubelet 

systemctl start --now kubelet 
systemctl status kubelet 

 

 

2. The cluster is broken again. Investigate and fix the issue.

 

k get nodes
ssh node01

systemctl status docker
systemctl status kubelet 

cat /etc/kubernetes/kubelet.conf

 : 이상없음

cat /var/lib/kubelet/config.yaml

: config 에 명시된 file 이름과 해당경로 파일 이름이 다르다. 

k edit /var/lib/kubelet/config.yaml

 : ca 이름명 변경

service kubelet restart
service kubelet status

 

3. The cluster is broken again. Investigate and fix the issue.

vi /etc/kubernetes/kubelet.conf

service kubelet restart
service kubelet status