그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그 그

3 minute read

Organizing Access Using kubeconfig Files

The kubectl command-line tool uses kubeconfig files to find the information it needs to choose a cluster and communicate with the API server of a cluster.

Problem

If you’ve become aware of a security breach that affects you, you may want to revoke or cycle credentials in case anything was leaked. However, this is not possible with the initial or master kubeconfig from your cluster.

teaser

Pitfall

Never distribute the kubeconfig, which you can download directly within the Gardener dashboard, for a productive cluster.

kubeconfig-dont

Create a Custom kubeconfig File for Each User

Create a separate kubeconfig for each user. One of the big advantages is, that you can revoke them and control the permissions better. A limitation to single namespaces is also possible here.

The script creates a new ServiceAccount with read privileges in the whole cluster (Secrets are excluded). To run the script jq, a lightweight and flexible command-line JSON processor, must be installed.

#!/bin/bash

if [[ -z "$1" ]] ;then
  echo "usage: $0 <username>"
  exit 1
fi

user=$1
kubectl create sa ${user}
secret=$(kubectl get sa ${user} -o json | jq -r .secrets[].name)
kubectl get secret ${secret} -o json | jq -r '.data["ca.crt"]' | base64 -D > ca.crt

user_token=$(kubectl get secret ${secret} -o json | jq -r '.data["token"]' | base64 -D)
c=`kubectl config current-context`
cluster_name=`kubectl config get-contexts $c | awk '{print $3}' | tail -n 1`
endpoint=`kubectl config view -o jsonpath="{.clusters[?(@.name == \"${cluster_name}\")].cluster.server}"`

# Set up the config
KUBECONFIG=k8s-${user}-conf kubectl config set-cluster ${cluster_name} \
    --embed-certs=true \
    --server=${endpoint} \
    --certificate-authority=./ca.crt

KUBECONFIG=k8s-${user}-conf kubectl config set-credentials ${user}-${cluster_name#cluster-} --token=${user_token}
KUBECONFIG=k8s-${user}-conf kubectl config set-context ${user}-${cluster_name#cluster-} \
    --cluster=${cluster_name} \
    --user=${user}-${cluster_name#cluster-}
KUBECONFIG=k8s-${user}-conf kubectl config use-context ${user}-${cluster_name#cluster-}

cat <<EOF | kubectl create -f -
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
  name: view-${user}-global
subjects:
- kind: ServiceAccount
  name: ${user}
  namespace: default
roleRef:
  kind: ClusterRole
  name: view
  apiGroup: rbac.authorization.k8s.io

EOF


echo "done! Test with: "
echo "export KUBECONFIG=k8s-${user}-conf"
echo "kubectl get pods"

If edit or admin rights are to be assigned, the ClusterRoleBinding must be adapted in the roleRef section with the roles listed below.

Furthermore, you can restrict this to a single namespace by not creating a ClusterRoleBinding but only a RoleBinding within the desired namespace.

Default ClusterRole	Default ClusterRoleBinding	Description
cluster-admin	system:masters group	Allows super-user access to perform any action on any resource. When used in a ClusterRoleBinding, it gives full control over every resource in the cluster and in all namespaces. When used in a RoleBinding, it gives full control over every resource in the rolebinding’s namespace, including the namespace itself.
admin	None	Allows admin access, intended to be granted within a namespace using a RoleBinding. If used in a RoleBinding, allows read/write access to most resources in a namespace, including the ability to create roles and rolebindings within the namespace. It does not allow write access to resource quota or to the namespace itself.
edit	None	Allows read/write access to most objects in a namespace. It does not allow viewing or modifying roles or rolebindings.
view	None	Allows read-only access to see most objects in a namespace. It does not allow viewing roles or rolebindings. It does not allow viewing secrets, since those are escalating.