Asking for help? Comment out what you need so we can get more information to help you!
Cluster information:
Kubernetes version: 1.20.0
Cloud being used: AWS
Installation method: GitHub - billchen8888/k8s-ansible at test_v1.20.0 (kube-apiserver is deployde as an OS level service)
Host OS: CentOS 8
CNI and version: cni_plugins_version: 0.8.6
CRI and version: docker 20.10.1
The issue I run into is:
kube-apiserver service is started. It only listen on port 6443, but not localhost:8080.
and this causes the following command fails when it is run on the master node:
kubectl create clusterrolebinding kubelet-bootstrap --clusterrole=system:node-bootstrapper --user=kubelet-bootstrap
error: failed to create clusterrolebinding: Post “http://localhost:8080/apis/rbac.authorization.k8s.io/v1/clusterrolebindings?fieldManager=kubectl-create”: dial tcp [::1]:8080: connect: connection refused
I have been use the exact same procedure to deploy k8s v1.18.3 and it works fine. The difference I noticed is that the kube-apiserver service in k8s v1.18.3 listens to localhost:8080
My question: do I miss anything in k8s v1.12.0 cofiguration? or kube-apiserver doesn’t list to localhost:8080 any more?