Scalable Jenkins on Kubernetes Cluster & Pipeline for Automated CI CD (Jenkins - Github Integration)

Setting up scalable Jenkins on Kubernetes can be a daunting task. This article demonstrates not only How to setup Scalable Jenkins on Kubernetes but also how to integrate Jenkins into Github so that whenever any Pull Request is requested on a branch, the build is automatically triggered and results are made available to the branch admin to review before merging the Pull Request.

Let us first get the scalable Jenkins up before setting up a pipeline for automated CI CD.

Before we go through the steps, here is the brief on the Master and Slave node docker images

Jenkins Master Node

Jenkins master uses the image brainupgrade/jenkins:2.274jdk11x

This image has many plugins required for smooth CD on Kubernetes. To know the image content, Dockerfile (./master/Dockerfile) is kept here. Following plugins are made available as part of this image to save time:

1. ssh-slaves
2. email-ext
3. mailer
4. slack
5. htmlpublisher
6. greenballs
7. simple-theme-plugin
8. kubernetes
9. workflow-aggregator
10. blueocean
11. sonar
12. maven-plugin

Jenkins Slave Node

Jenkins slave uses the image brainupgrade/jnlp-slave:1.0.0

This image is based on openjdk11 containing maven, docker runtime so that spring boot project can be checked out, maven built including docker image building & deployment.

Steps — Scalable Jenkins on Kubernetes Cluster

Below is the step-by-step guide to making it happen.

Launch Jenkins master

Run below Kubernetes configurations

kubectl create ns jenkins
kubectl apply -f rbac.yaml
kubectl apply -f deploy.yaml
kubectl apply -f service.yaml

Secure the master

Once Jenkins master server is deployed, it would take few minutes to get the UI up and running To access UI, run below

kubectl port-forward deploy/jenkins 8080:8080

and launch URL https://localhost:8080 on the browser

Go to https://localhost:8080/configureSecurity/ and secure the server by enabling security as shown in the below picture.

Once you save it, you will get an option to set a username and password.

After login, come back to this URL again and select tickmark to Agent — Controller Security. This option gets visible after setting up a username and password.

Configure the Kubernetes plugin as shown in the snapshots below

Now go to the URL: https://localhost:8080/configureClouds and key in the configuration as shown in the below snapshots.

Setup Global credentials

To test the docker commands especially login, first set up the global credentials as shown in below

Create Build Job

Create docker-test job as Pipeline, click OK and insert below text in the pipeline block

pipeline {
    agent {
        kubernetes{
            label 'jenkins-slave'
        }
    }
    environment{
        DOCKER_USERNAME = 'brainupgrade'
        DOCKER_PASSWORD = credentials('docker-brainupgrade')
    }
    stages {
        stage('docker login') {
            steps{
                sh(script: """
                    docker login -u $DOCKER_USERNAME -p $DOCKER_PASSWORD
                """, returnStdout: true) 
            }
        }
    }
}

Click SAVE

Test the pipeline

Run the docker-test job and you would see that a pod will be launched by Jenkins master to run the docker-test build job and the pod will be terminated immediately build completes.

Once the build job completes, the build pod is terminated

Steps — Jenkins Pipeline for Automated CI CD

To illustrate, automated CI CD, I have setup an example github project Weather Service Also created a branch named jenkins and whenever any contributions are made to this branch via pull request, branch / project admin can review the pull request, view the build and test status before merging into jenkins branch

Jenkins Job

On Jenkins, setup a build job named weather-service and type pipeline, click OK and then scroll down to insert content of this pipeline file into the pipeline section.

Install Plugin

Also, install the below plugin.

1. GitHub pull request builder

Github Setup

Now generate an access token using https://github.com/settings/tokens/new (select repo:status)

Create a webhook using your Github repo settings

1. Insert https://<Your_Jenkins_Public_URL>/ghprbhook/
2. Event trigger — Select individual events (Pull requests, Issue comments) So that whenever any pull request is created, Github can notify your Jenkins URL

Jenkins Setup

Create Jenkins Credentials (type: Secret text) using the above access token say github-bu-token

Configure GitHub Server here https://localhost:8080/configure

Scroll down and configure Github Pull Request Builder and choose the right credentials

Create a new job say weather-service, select pipeline, and click OK

Select Github Pull Request Builder, see the below snapshot

Click on advanced and insert whitelist target branch say Jenkins. You should also select Build every pull request automatically without asking to ensure that every pull request is good enough in terms of build and test quality before merging

Test the setup

Now, you can raise a pull request on the selected branch (Jenkins in my case) and you would notice that the build gets triggered automatically and its results will be shown on the Pull request page so that the branch admin can review the outcome and decide on the merge.

Build Status

To show the build status on your main Github page, install the plugin embeddable-build-status and goto your build job and click on this plugin link to generate a URL that would look similar to the one below

[![Build Status](https://<Your_Jenkins_Public_URL>/buildStatus/icon?job=<job_name>)](https://<Your_Jenkins_Public_URL>/job/<job_name>/)

Kubernetes Cluster Info

Below is the info on the tool used to setup the Kubernetes cluster and its version

1. Kubernetes Server 1.19.7
2. Cluster Management Tool — kops
3. Cloud Provider AWS

Conclusion

This article demonstrated How to setup Jenkins Pipeline to achieve automated CI CD including Github and Jenkins integration. The moment Git Pull request is raised on a particular branch, build is triggered and results are made available to the branch admin to review before merging the pull request.

About the Author

The author, Rajesh G, is The Chief Architect @ Brain Upgrade Academy where he has designed the IoT-based Fleet Management Platform that runs on a Kubernetes Cluster on AWS Amazon. He is also a certified Kubernetes Administrator and TOGAF certified Enterprise Architect. Rajesh led various digital transformation initiatives for Fortune 500 FinTech companies. Over the last 20+ years, he has been part of many successful technology startups.

About Brain Upgrade Academy

We, at Brain Upgrade, offer Kubernetes Consulting services to our clients including Up Skilling (training) of clients teams thus facilitate efficient utilization of Kubernetes Platform. To know more on the Kubernetes please visit www.brainupgrade.in/blog and register on www.brainupgrade.in/enroll to equip yourself with Kubernetes skills.

Why Brain Upgrade

We at Brain Upgrade, partner with our customers in the digital transformation of their businesses by providing:

Technology Consulting in product development, IoT, DevOps, Cloud, Containerization, Big Data Analysis with a heavy focus on Open source technologies. Training the IT workforce on the latest cloud technologies such as Kubernetes, Docker, AI, ML, etc.

You may want to register for the upcoming trainings on https://brainupgrade.in/enroll