Skip to content

Latest commit

 

History

History
24 lines (17 loc) · 1.23 KB

kube-redeploy-coredns.md

File metadata and controls

24 lines (17 loc) · 1.23 KB

Redeploy CoreDNS Services

Due to a bug / feature of kubeadm, if coredns is deployed without additional nodes available (e.g. another master or worker nodes), then it will deploy entirely on the same node (e.g. master01).

NAME                                   READY   STATUS    RESTARTS   AGE     IP             NODE           NOMINATED NODE   READINESS GATES
coredns-5644d7b6d9-jdrbt               1/1     Running   1          2d16h   10.32.0.2      master01   <none>           <none>
coredns-5644d7b6d9-rsxwf               1/1     Running   1          2d16h   10.32.0.3      master01   <none>           <none>

To resolve this, we can restart the deployment of coredns, which will deploy across different nodes to provide the HA level of redundancy we've come to expect from Kubernetes.

master01

kubectl -n kube-system rollout restart deployment coredns
kubectl get pods -A -o wide | grep coredns

kube-system            coredns-58ddcb86c5-cxlnl                      1/1     Running   0          12m     10.42.0.4      worker03   <none>           <none>
kube-system            coredns-58ddcb86c5-dqzh8                      1/1     Running   0          12m     10.36.0.5      worker01   <none>           <none>