post image :date_long | 2 min Read

CKS Restrict API server

There is an flag when starting kube-aoiserver called:

kube-apiserver --anonymous-auth=true|false

The default value for this option is true because some liveness and readiness probes needs it.

Then we have –insecure-port=8080 (deprecated)

There was an option in past before Kubernetes 1.20 to use In such a case, all authentication and authorization to API server would be skipped and this was and still is considered as highly insecure and dangerous.

kube-apiserver --insecure-port=8080

Call API server via curl

# Certification authority
k config view -o jsonpath='{.clusters[].cluster.certificate-authority-data}' --raw | base64 -d > ca

# Client certificate
k config view -o jsonpath='{.users[].user.client-certificate-data}' --raw  | base64 -d > crt

# Client key
k config view -o jsonpath='{.users[].user.client-key-data}' --raw | base64 -d > key

curl https://127.0.0.1:6443 --cacert ca --cert crt --key key
{
  "paths": [
    "/.well-known/openid-configuration",
    "/api",
    "/api/v1",
    "/apis",
    "/apis/",
    "/apis/admissionregistration.k8s.io",
    "/apis/admissionregistration.k8s.io/v1",
    "/apis/apiextensions.k8s.io",
    "/apis/apiextensions.k8s.io/v1",
    "/apis/apiregistration.k8s.io",
    "/apis/apiregistration.k8s.io/v1",
    "/apis/apps",
    "/apis/apps/v1",
    "/apis/authentication.k8s.io",
    "/apis/authentication.k8s.io/v1",
    "/apis/authorization.k8s.io",
    "/apis/authorization.k8s.io/v1",
    "/apis/autoscaling",
    "/apis/autoscaling/v1",
    "/apis/autoscaling/v2"
    ...
}

NodeRestriction plugin

Kubernetes nodes do have their own kubeconfig file located at /etc/kubernetes/kubelet.conf to communicate to kube-apiserver however, their permissions are limited.

    - kube-apiserver
    ...
    - --enable-admission-plugins=NodeRestriction
    ...

Here are several examples what can/cannot be done by using this KUBECONFIG file

root@lima-k8s:~# kubectl get pods --kubeconfig /etc/kubernetes/kubelet.conf
NAME             READY   STATUS    RESTARTS   AGE
accessor         1/1     Running   0          26h
krissko          1/1     Running   0          6d1h
pod-in-default   1/1     Running   0          6d1h
root@lima-k8s:~# kubectl get nodes --kubeconfig /etc/kubernetes/kubelet.conf
NAME       STATUS   ROLES                  AGE    VERSION
lima-k8s   Ready    control-plane,master   6d1h   v1.23.4
root@lima-k8s:~# kubectl get ns --kubeconfig /etc/kubernetes/kubelet.conf
Error from server (Forbidden): namespaces is forbidden: User "system:node:lima-k8s" cannot list resource "namespaces" in API group "" at the cluster scope

We are not allowed to modify special labels like

root@lima-k8s:~# kubectl label node lima-k8s node-restriction.kubernetes.io/test=yes
Error from server (Forbidden): nodes "lima-k8s" is forbidden: is not allowed to modify labels: node-restriction.kubernetes.io/test```


#### Troubleshooting


crictl ps -a CONTAINER IMAGE CREATED STATE NAME ATTEMPT POD ID 9b69ae66f21ad b6d7abedde399 25 seconds ago Exited kube-apiserver 4 0bbfc476e406a … controlplane $ crictl logs -f 9b69ae66f21ad Error: unknown flag: –this-is-very-wrong



###### Log locations to check:

/var/log/pods /var/log/containers crictl ps + crictl logs docker ps + docker logs (in case when Docker is used) kubelet logs: /var/log/syslog or journalctl


###### NodeRestriction in kube-apiserver manifest

We need to enable the NodeRestriction in the Apiserver manifest

spec: containers:

  • command:
    • kube-apiserver
    • –advertise-address=172.30.1.2
    • –allow-privileged=true
    • –authorization-mode=Node,RBAC
    • –client-ca-file=/etc/kubernetes/pki/ca.crt
    • –enable-admission-plugins=NodeRestriction
    • –enable-bootstrap-token-auth=true
author image

Jan Toth

I have been in DevOps related jobs for past 6 years dealing mainly with Kubernetes in AWS and on-premise as well. I spent quite a lot …

comments powered by Disqus