Setup a Kubernetes Cluster

In this module, you set up a Kubernetes cluster that has Istio installed and a namespace to use throughout the tutorial.

  1. Ensure you have access to a Kubernetes cluster. You can use the Google Kubernetes Engine or the IBM Cloud Kubernetes Service.

  2. Create an environment variable to store the name of a namespace that you will use when you run the tutorial commands. You can use any name, for example tutorial.

    $ export NAMESPACE=tutorial
    
  3. Create the namespace:

    $ kubectl create namespace $NAMESPACE
    
  4. Install Istio.

  5. Enable Envoy’s access logging.

  6. Create a Kubernetes Ingress resource for these common Istio services using the kubectl command shown. It is not necessary to be familiar with each of these services at this point in the tutorial.

    The kubectl command can accept an in-line configuration to create the Ingress resources for each service:

    $ kubectl apply -f - <<EOF
    apiVersion: extensions/v1beta1
    kind: Ingress
    metadata:
      name: istio-system
      namespace: istio-system
    spec:
      rules:
      - host: my-istio-dashboard.io
        http:
          paths:
          - path: /
            backend:
              serviceName: grafana
              servicePort: 3000
      - host: my-istio-tracing.io
        http:
          paths:
          - path: /
            backend:
              serviceName: tracing
              servicePort: 9411
      - host: my-istio-logs-database.io
        http:
          paths:
          - path: /
            backend:
              serviceName: prometheus
              servicePort: 9090
      - host: my-kiali.io
        http:
          paths:
          - path: /
            backend:
              serviceName: kiali
              servicePort: 20001
    EOF
    
  7. Create a role to provide read access to the istio-system namespace. This role is required to limit permissions of the participants in the steps below.

    $ kubectl apply -f - <<EOF
    kind: Role
    apiVersion: rbac.authorization.k8s.io/v1beta1
    metadata:
      name: istio-system-access
      namespace: istio-system
    rules:
    - apiGroups: ["", "extensions", "apps"]
      resources: ["*"]
      verbs: ["get", "list"]
    EOF
    
  8. Create a service account for each participant:

    $ kubectl apply -f - <<EOF
    apiVersion: v1
    kind: ServiceAccount
    metadata:
      name: ${NAMESPACE}-user
      namespace: $NAMESPACE
    EOF
    
  9. Limit each participant’s permissions. During the tutorial, participants only need to create resources in their namespace and to read resources from istio-system namespace. It is a good practice, even if using your own cluster, to avoid interfering with other namespaces in your cluster.

    Create a role to allow read-write access to each participant’s namespace. Bind the participant’s service account to this role and to the role for reading resources from istio-system:

    $ kubectl apply -f - <<EOF
    kind: Role
    apiVersion: rbac.authorization.k8s.io/v1beta1
    metadata:
      name: ${NAMESPACE}-access
      namespace: $NAMESPACE
    rules:
    - apiGroups: ["", "extensions", "apps", "networking.k8s.io", "networking.istio.io", "authentication.istio.io",
                  "rbac.istio.io", "config.istio.io"]
      resources: ["*"]
      verbs: ["*"]
    ---
    kind: RoleBinding
    apiVersion: rbac.authorization.k8s.io/v1beta1
    metadata:
      name: ${NAMESPACE}-access
      namespace: $NAMESPACE
    subjects:
    - kind: ServiceAccount
      name: ${NAMESPACE}-user
      namespace: $NAMESPACE
    roleRef:
      apiGroup: rbac.authorization.k8s.io
      kind: Role
      name: ${NAMESPACE}-access
    ---
    kind: RoleBinding
    apiVersion: rbac.authorization.k8s.io/v1beta1
    metadata:
      name: ${NAMESPACE}-istio-system-access
      namespace: istio-system
    subjects:
    - kind: ServiceAccount
      name: ${NAMESPACE}-user
      namespace: $NAMESPACE
    roleRef:
      apiGroup: rbac.authorization.k8s.io
      kind: Role
      name: istio-system-access
    EOF
    
  10. Each participant needs to use their own Kubernetes configuration file. This configuration file specifies the cluster details, the service account, the credentials and the namespace of the participant. The kubectl command uses the configuration file to operate on the cluster.

    Generate a Kubernetes configuration file for each participant:

    $ cat <<EOF > ./${NAMESPACE}-user-config.yaml
    apiVersion: v1
    kind: Config
    preferences: {}
    
    clusters:
    - cluster:
        certificate-authority-data: $(kubectl get secret $(kubectl get sa ${NAMESPACE}-user -n $NAMESPACE -o jsonpath={.secrets..name}) -n $NAMESPACE -o jsonpath='{.data.ca\.crt}')
        server: $(kubectl config view -o jsonpath="{.clusters[?(.name==\"$(kubectl config view -o jsonpath="{.contexts[?(.name==\"$(kubectl config current-context)\")].context.cluster}")\")].cluster.server}")
      name: ${NAMESPACE}-cluster
    
    users:
    - name: ${NAMESPACE}-user
      user:
        as-user-extra: {}
        client-key-data: $(kubectl get secret $(kubectl get sa ${NAMESPACE}-user -n $NAMESPACE -o jsonpath={.secrets..name}) -n $NAMESPACE -o jsonpath='{.data.ca\.crt}')
        token: $(kubectl get secret $(kubectl get sa ${NAMESPACE}-user -n $NAMESPACE -o jsonpath={.secrets..name}) -n $NAMESPACE -o jsonpath={.data.token} | base64 --decode)
    
    contexts:
    - context:
        cluster: ${NAMESPACE}-cluster
        namespace: ${NAMESPACE}
        user: ${NAMESPACE}-user
      name: ${NAMESPACE}
    
    current-context: ${NAMESPACE}
    EOF
    
  11. Set the KUBECONFIG environment variable for the ${NAMESPACE}-user-config.yaml configuration file:

    $ export KUBECONFIG=./${NAMESPACE}-user-config.yaml
    
  12. Verify that the configuration took effect by printing the current namespace:

    $ kubectl config view -o jsonpath="{.contexts[?(@.name==\"$(kubectl config current-context)\")].context.namespace}"
    tutorial
    

    You should see the name of your namespace in the output.

  13. If you are setting up the cluster for yourself, copy the ${NAMESPACE}-user-config.yaml file mentioned in the previous steps to your local computer, where ${NAMESPACE} is the name of the namespace you provided in the previous steps. For example, tutorial-user-config.yaml. You will need this file later in the tutorial.

    If you are an instructor, send the generated configuration files to each participant. The participants must copy their configuration file to their local computer.

Congratulations, you configured your cluster for the tutorial!

You are ready to setup a local computer.

Esta informação foi útil?
Você tem alguma sugestão de melhoria?

Obrigado pelo seu feedback!