SETUP-ON-GKE.MD

Setup on GKE

Prerequisites

1. Create GCP Account

Install the Cluster

1. Sign-in to Google Cloud Platform Console

2. Install and initialize the Google Cloud SDK

   Note: When you initialize the Google Cloud SDK, be sure to Create a new project named daytrader

   $ gcloud container clusters create daytrader-gke-cluster

   NAME	LOCATION	MASTER_VERSION	MASTER_IP	MACHINE_TYPE	NODE_VERSION	NUM_NODES	STATUS
   gke-daytrader-cluster	us-central1-a	1.9.7-gke.11	xx.xx.xx.xx	n1-standard-1	1.9.7-gke.11	3	RUNNING

3. Get authentication credentials for the cluster

   $ gcloud container clusters get-credentials daytrader-gke-cluster

   This command updates your kubeconfig so that the kubectl command can connect to your new cluster.

4. Configure kubeconfig

At this point you should have a cluster and ~/.kube/config file created. That file has information about the cluster(s) you created. The kubectl command line interface uses that information to locate and authenticate to a cluster. If you have multiple clusters, then you can use kubectl config use-context CONTEXT_NAME command to point kubectl to a specific cluster. To make it easier for you to work with multiple clusters, edit the context name in the config file. For example,

Original ~/.kube/config

- context:
    cluster: gke_daytrader-xxxxxx_us-central1-a_daytrader-gke-cluster
    user: gke_daytrader-xxxxxx_us-central1-a_daytrader-gke-cluster
  name: gke_daytrader-xxxxxx_us-central1-a_daytrader-gke-cluster

Modified ~/.kube/config

- context:
    cluster: gke_daytrader-xxxxxx_us-central1-a_daytrader-gke-cluster
    user: gke_daytrader-xxxxxx_us-central1-a_daytrader-gke-cluster
  name: daytrader-gke-cluster

5. Verify that the cluster is running

   $ kubectl cluster-info

   You should see Kubernetes master is running at https://35.184.21.226

6. Verify that the worker nodes are running

   a. $ kubectl config use-context daytrader-gke-cluster

   You should see Switched to context daytrader-gke-cluster

   b. $ kubectl get nodes

   NAME	STATUS	ROLES	AGE	VERSION
   gke-daytrader-gke-cluste-default-pool-xxxxxxxx-xxxx	Ready	<none>	7m11s	v1.9.7-gke.11
   gke-daytrader-gke-cluste-default-pool-xxxxxxxx-xxxx	Ready	<none>	7m11s	v1.9.7-gke.11
   gke-daytrader-gke-cluste-default-pool-xxxxxxxx-xxxx	Ready	<none>	7m11s	v1.9.7-gke.11

Install the NGINX Ingress Controller

1. Create the cluster role binding

   $ kubectl create clusterrolebinding cluster-admin-binding --clusterrole cluster-admin --user $(gcloud config get-value account)

   This command creates a cluster role binding to administer the cluster and adds your GCP user account to that role

   Note: other Kubernetes platforms automatically add the user that created the cluster to the cluster-admin role.

2. Create the namespace and configmaps; otherwise cloud-generic.yaml and ingress controller will have errors

   a. $ cd daytrader-webapp/daytrader-web/env/external/kubernetes/conf

   b. $ kubectl apply -f ingress-nginx-namespace.yaml

   c. $ kubectl apply -f ingress-nginx-configmap.yaml

3. Deploy the NGINX Load Balancer

   a. $ kubectl apply -f https://raw.githubusercontent.com/kubernetes/ingress-nginx/master/deploy/provider/cloud-generic.yaml

   b. $ kubectl -n ingress-nginx get services

   NAME	TYPE	CLUSTER-IP	EXTERNAL-IP	PORT(S)	AGE
   ingress-nginx	LoadBalancer	10.7.252.186	xx.xx.xx.xx	80:31884/TCP,443:31058/TCP	55s

   Wait for the load balancer to be assigned an EXTERNAL-IP. It may take a few minutes

4. Deploy the NGINX Ingress Controller

   a. kubectl apply -f https://raw.githubusercontent.com/kubernetes/ingress-nginx/master/deploy/mandatory.yaml

   b. $ kubectl -n ingress-nginx get pods

   NAME	READY	STATUS	RESTARTS	AGE
   nginx-ingress-controller-xxxxxxxxx-xxxxx	1/1	Running	0	39s

   Wait for the controller start. It may take a few minutes

5. Review the logs to verify the installation. This is also useful for troubleshooting ingress resources.

   $ kubectl -n ingress-nginx logs nginx-ingress-controller-xxxxxxxxx-xxxxx

   You should see successfully acquired lease ingress-nginx/ingress-controller-leader-nginx

Install the Helm Server (Tiller)

To install tiller into the gke cluster

1. $ kubectl use-context CLUSTER_NAME

2. Create a service account for tiller

   $ kubectl -n kube-system create serviceaccount tiller

3. Create a cluster role binding for tiller

   $ kubectl create clusterrolebinding tiller --clusterrole cluster-admin --serviceaccount=kube-system:tiller

4. Install helm within the cluster

   $ helm init --service-account tiller

5. Verify the installation

   $ kubectl -n kube-system get deployments,services tiller-deploy

   NAME	DESIRED	CURRENT	UP-TO-DATE	AVAILABLE	AGE
   deployment.extensions/tiller-deploy	1	1	1	1	39s

   NAME	TYPE	CLUSTER-IP	EXTERNAL-IP	PORT(S)	AGE
   service/tiller-deploy	ClusterIP	xx.xx.xx.xx	none	xxxxx/TCP	39s

Notes: in this installation, Tiller is deployed with an insecure allow unauthenticted users policy. This is completely appropriate for local development or for single tenant clusters within a private network. To apply security configurations for multi-tenant clusters in uncontrolled environments, or for clusters with access to high valued data, see Securing your Helm Installation

Install the Kubernetes Dashboard (N/A)

The Kubernetes Dashboard is disabled and deprecated in GKE. instead, use the GKE Dashboard through the Console.

Resources

1. Google Kubernetes Engine