Cyberithub

Step by Step Guide to Create a GKE Cluster in Google Cloud

Advertisements

GKE is a google cloud service which provides managed environment to deploy, manage and scale containerized applications. It becomes easier to create a cluster using GKE with required number of nodes with just few clicks. Nodes in the cluster are VM instances which can be created using another service called Compute Engine. In this tutorial, We will create a cluster, deploy a microservice and play with it using kubectl utility.

Step by Step Guide to Create a GKE Cluster in Google Cloud

Step by Step Guide to Create a GKE Cluster in Google Cloud

Create a Kubernetes cluster with the default node pool using below steps:-

  • Click on "My First Project"
  • Click on "New Project"
  • Name the project as "first-demo" (you can give any name )
  • Click on "Create"

Project is created successfully. Now open the project and copy the project id for further steps. But before going any further, we first need to enable the Kubernetes service and create cluster. To do so, follow below steps:-

  • Go to project "first-demo".
  • Search service "Kubernetes engine".
  • Click "enable" to enable Kubernetes Engine API.
  • Once enabled, click "Create" to create the cluster.
  • Create the cluster with all default settings and name the cluster as "my-first-cluster".

Cluster is created successfully with 3 default nodes, 6 CPUs(2 for each node) and 12GB memory.

NOTE:

In GCP, we always first enable the service API before using any service. For example to use GKE, enable Kubernetes Engine API first.

Step 1: Login to cloud shell

Next we will login to cloud shell. To do so, go to cloud shell and type below command.

gcloud config set project first-demo-311705

Step 2: Connect to the Kubernetes cluster

Now that the cluster is create, we will connect to the cluster for deploying microservice. To do so, follow below steps:-

Advertisements
  • Go to project "first-demo". Click on 3 dots.
  • Click on "connect". copy the command which pop up shows.
  • Go back to cloud shell and type below command.
cyberithub@cloudshell:~ (first-demo-311705)$ gcloud container clusters get-credentials my-first-cluster --zone us-central1-c --project first-demo-311705
Fetching cluster endpoint and auth data.
kubeconfig entry generated for my-first-cluster.

Where,
my-first-cluster is the cluster name.
us-central1-c is the zone name.
first-demo-311705 is the project name.

When a cluster is created, certain directories also gets created which stores the cluster information.

cyberithub@cloudshell:~ (first-demo-311705)$ ls -la
total 40
drwxr-xr-x  5  cyberithub cyberithub 4096  Apr 24 06:38 .
drwxr-xr-x  4  root       root       4096  Apr 23 11:49 ..
-rw-------  1  cyberithub cyberithub 2810  Apr 24 18:06 .bash_history
-rw-r--r--  1  cyberithub cyberithub 220   Apr 18 2019  .bash_logout
-rw-r--r--  1  cyberithub cyberithub 3564  Apr 17 07:24 .bashrc
drwxr-xr-x  3  cyberithub cyberithub 4096  Apr 17 07:04 .config
drwxr-xr-x  2  cyberithub cyberithub 4096  Apr 23 11:49 .docker
drwxr-xr-x  3  cyberithub cyberithub 4096  Apr 24 18:05 .kube
-rw-r--r--  1  cyberithub cyberithub 807   Apr 18 2019  .profile
-rw-r--r--  1  cyberithub cyberithub 913   Apr 24 17:45  README-cloudshell.txt

Kube config file is stored on path /home/ankur123sh/.kube/config .This config has the information of all the clusters created. Open this file and check the content to get more insight.

cyberithub@cloudshell:~/.kube (first-demo-311705)$ ls -lhtr
total 8.0K
drwxr-x--- 4 cyberithub cyberithub 4.0K Apr 24 14:40 cache
-rw------- 1 cyberithub cyberithub 2.7K Apr 24 18:05 config

Step 3: Deploy microservice to Kubernetes

Let's create deployment using kubectl utility. I am using one of the image from docker hub which can be referred on Docker Hub website.

cyberithub@cloudshell:~ (first-demo-311705)$ kubectl create deployment first-deployment --image=openwhisk/python2action:nightly
deployment.apps/first-deployment created

where,
first-deployment is the deployment name.
openwhisk/python2action:nightly is the docker image (openwhisk -> docker id, python2action ->image name, nightly image tag we want to use ).

Command to view running deployment:-

cyberithub@cloudshell:~ (first-demo-311705)$ kubectl get deployment
NAME                READY UP-TO-DATE AVAILABLE  AGE
first-deployment    1/1      1           1      117s

Next we will expose the deployment to outside world by using below command. What we get from exposing deployment to outside world is a service. The type of service we are creating here is LoadBalancer.

cyberithub@cloudshell:~ (first-demo-311705)$ kubectl expose deployment first-deployment --type=LoadBalancer --port=8080
service/first-deployment exposed

Below is the command to view running services:-

cyberithub@cloudshell:~ (first-demo-311705)$ kubectl get service
NAME               TYPE           CLUSTER-IP  EXTERNAL-IP        PORT(S)            AGE
first-deployment LoadBalancer    10.76.7.140  35.222.153.200   8080:31910/TCP     4m15s
kubernetes       ClusterIP        10.76.0.1    <none>           443/TCP            8h

Increase number of instances of microservice using below command. We will create 4 instances of our microservice using flag replicas.

cyberithub@cloudshell:~ (first-demo-311705)$ kubectl scale deployment first-deployment --replicas=4
deployment.apps/first-deployment scaled

cyberithub@cloudshell:~ (first-demo-311705)$ kubectl get deployment
      NAME          READY      UP-TO-DATE  AVAILABLE   AGE
first-deployment     4/4             4        4        36m

To see the status of each instances(also called as pod) use below command:-

cyberithub@cloudshell:~ (first-demo-311705)$ kubectl get pods
                    NAME           READY    STATUS    RESTARTS      AGE
first-deployment-87f979d87-72f98   1/1      Running      0          36m
first-deployment-87f979d87-gjdfj   1/1      Running      0          44s
first-deployment-87f979d87-n4w46   1/1      Running      0          44s
first-deployment-87f979d87-sng8z   1/1      Running      0          44s

Step 4: Increase/Decrease number of nodes in the cluster using gcloud utility

We can also scale the nodes used in our cluster based on usage. If the no of pods increases, it will require more no of resources. Let's resize the node pool to 2 using below command.

cyberithub@cloudshell:~ (first-demo-311705)$ gcloud container clusters resize my-first-cluster --node-pool default-pool --num-nodes=2 --zone=us-central1-c
Pool [default-pool] for [my-first-cluster] will be resized to 2.
Do you want to continue (Y/n)? y
Resizing my-first-cluster...done.
Updated [https://container.googleapis.com/v1/projects/first-demo-311705/zones/us-central1-c/clusters/my-first-cluster].

Where,
my-first-cluster is the cluster name.
default-pool is the Node pool name.
us-central1-c is the zone name.

Step 5: Setup auto scaling for our microservice

Instead of scaling microservice or resources every time, we can use auto-scaling for same job. Let's auto-scale microservice using below command:-

cyberithub@cloudshell:~ (first-demo-311705)$ kubectl autoscale deployment first-deployment --max=5 --cpu-percent=60
horizontalpodautoscaler.autoscaling/first-deployment autoscaled

cyberithub@cloudshell:~ (first-demo-311705)$ kubectl get hpa
       NAME              REFERENCE              TARGETS           MINPODS      MAXPODS REPLICAS AGE
first-deployment Deployment/first-deployment  <unknown>/60%        1               5      3     19m

Step 6: Setup auto scaling for our cluster

Let's auto-scale cluster using below command:-

cyberithub@cloudshell:~ (first-demo-311705)$ gcloud container clusters update my-first-cluster --enable-autoscaling --zone=us-central1-c --min-nodes=1 --max-nodes=6
Updating my-first-cluster...done.
Updated [https://container.googleapis.com/v1/projects/first-demo-311705/zones/us-central1-c/clusters/my-first-cluster].
To inspect the contents of your cluster, go to:
https://console.cloud.google.com/kubernetes/workload_/gcloud/us-central1-c/my-first-cluster?project=first-demo-311705

Step 7: To display cluster info

If you want to check the cluster info then you need to use kubectl cluster-info command as shown below.

cyberithub@cloudshell:~ (first-demo-311705)$ kubectl cluster-info
Kubernetes control plane is running at https://34.122.170.70
GLBCDefaultBackend is running at https://34.122.170.70/api/v1/namespaces/kube-system/services/default-http-backend:http/proxy
KubeDNS is running at https://34.122.170.70/api/v1/namespaces/kube-system/services/kube-dns:dns/proxy
Metrics-server is running at https://34.122.170.70/api/v1/namespaces/kube-system/services/https:metrics-server:/proxy

To further debug and diagnose cluster problems, use 'kubectl cluster-info dump'.

Step 8: To display current working context

If you want to check the current working context then you need to use kubectl config current-context command as shown below.

cyberithub@cloudshell:~ (first-demo-311705)$ kubectl config current-context
gke_first-demo-311705_us-central1-c_my-first-cluster

Leave a Comment