Recently, We did upgrade of client Kubernetes cluster and find that all pods started with IPs from 172.17.0.0/24. The customer has many pods that caused issue that pods use the same IPs on different nodes and We had to fix it.
To fix that issue We added following lines to the /etc/kubernetes/manifest/kube-controller-manager.conf :
- --allocate-node-cidrs=true - --cluster-cidr="10.1.0.0/16"
Following change cause kube-controller-manager pod restart.
Also, We did restart of all pods and it’s help to resolve this issue.
You can find more Kubernetes guides by link https://linuxnotes.org/category/kubernetes-guides/
Viva La Linux!