Jenkin_on_GKE

You now have access to Jenkins Serviceand a Kubernetes cluster managed by GKE. To take this solution further, you could use these components in your CD pipeline.

Core Steps:

(1) create GKE cluster

(2) install Helm CLI

(3) install Jenkins and connect to Jenkins Server

---

GKE

from step 1:

Activate Cloud Shell

• 1.1, setup GCE zone.

  gcloud config set compute/zone <zone>
  

• 1.2, create GKE cluster.

  gcloud container clusters create j-ci-cd \
   --machine-type <machine-type> \
   --cluster-version <version> \
   --scopes "https://www.googleapis.com/auth/source.read_write,cloud-platform"
   // scopes enable Jenkins to access container registry and cloud source repository.
  

• 1.3, check cluster is running.

  gcloud container clusters list
  
  [output]
  Name      location   master_version    master_ip   machine_type    node_version   num_node status
  j-ci-cd     <zone>    <version>      35.231.8.57   <machine type>    <version>    2        running
  

• 1.4, cfrm that connectivity to cluster

   kubectl cluster-info
   
   [output]
   kubernets master is running at <ip> 
   backend is running at <ip/api/v1/proxy/namesapces/kube-system/services/default-http-backend>
   kubeDNS is running at <ip/api/v1/proxy/namesapces/kube-system/services/kube-dns>
  

Helm

from step 2:

install and utile Helm CLI tool

• 2.1, download Helm binary.

  wget https://get.helm.sh/helm-v3.2.1-linux-amd64.tar.gz
  

• 2.2, unzip the Helm file.

  tar zxfv helm-v3.2.1-linux-amd64.tar.gz
  

• 2.3, copy the file to local system.

  cp linux-amd64/helm .
  

• 2.4, add role to RBAC for adding jenkins permissions afterward.

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

• 2.5, add repo

   ./helm repo add jenkins-ci https://charts.jenkins.io
   ./helm repo update 
   ./helm version
  

Jenkins

from step3:

install and util Jenkins

• 3.1, use helm cli to install jenkins chart with yaml file.

  ./helm install cd-jenkins -f jenkins/values.yaml jenkinsci/jenkins 
      --version 2.6.4
      --wait
  

• 3.2, check the Pod installing Jenkins is at running state

    kubectl get pods
    
    [output]
    Name                  ready   status     AGE
    cd-jenkins-<code>      1/1    running     1m
  

• 3.3, forward port from cloud shell to jenkins UI.

  export POD_NAME=$(kubectl get pods 
  --namespace default 
  -l "app.kubernetes.io/component=jenkins-master" 
  -l "app.kubernetes.io/instance=cd-jenkins" 
  -o jsonpath="{.items[0].metadata.name}")
  kubectl port-forward $POD_NAME 8080:8080 >> /dev/null &
  

• 3.4, check Jenkins Service is created well.

  kubectl get svc
  
  [output]
  Name               cluster-ip     external-ip    port(s)    gae
  cd-jenkins         <internal-ip1>   <none>      8080/tcp     3h
  cd-jenkins-agent   <internal-ip2>   <none>     50000/tcp     3h
  kubernets          <internal-ip3>   <none>       443/tcp     9h
  

• tips & attentions

  The Jenkins installation is using the Kubernetes Plugin to create builder agents. They will be automatically launched as necessary when the Jenkins master needs to run a build. When their work is done, they are automatically terminated and their resources are added back to the cluster's resource pool.

• 3.5, to retrieve the admins password and connect to Jenkins Server.

  printf $(kubectl get secret cd-jenkins 
          -o jsonpath="{.data.jenkins-admin-password}" 
          | base64 --decode)
          ;echo
  

• 3.6, to open Jenkins UI (chart), plz click on "Web Preview" tab.

• 3.7, log in by clicking on the "log in" tab, and enter the "admin" for user field, and password for the password filed, then click on "log in" button.

Ref Doc

https://cloud.google.com/solutions/jenkins-on-kubernetes-engine-tutorial