Project: Continuous Integration & Deployment using Cloudformation, Jenkins, Docker & Kubernetes (EKS)

Description

In this project continuous pipeline is created in order to demonstarte deployment & production stages.

The project is divided in three main parts:

• Creating a cloud netowrk and an EKS cluster in AWS with two nodes
• Code commits and pushed to docker hub
• Image pulled from docker registry and deployed to kubernetes pods

Understanding Project

• First step is where we implemented clouformation stack to implement VPC, subnets, security group, an EKS cluster and launch configuration with two nodes.

• Second step includes an independent EC2 instance with jenkins and docker installed and configured.

• Third step include kiubernetes configured with AWS EKS cluster in cloud

• Forth step is the pipeline itself that take the commited code from github and push it to docker repository

• The pipeline include pulling the image from docker registry and deployment into kuberenetes pods through a load balancer

• Target group is assigned and health checks are performed on both instances.

• Blue deployment is configured by default on three kubernetes pods

• After the new commit, code is passed to docker hub

• Code is pulled from docker registry and to a temperory green deployment; initiated on runtime.

• Green deployment creates three new pods in the pipeline. Total of 6 pods now in pipeline

• Blue deployment is deleted, deleting three pods.

• In the mean time temperory green deployment is attached to load balancer showing updated website

• New deployment is created with the name of Blue with the new updates.

• Temperory green deployment deleted to complete the circle and we are back on blue deployment with the updated website Notes * Project is configured as blue / green deployment in kubernetes pods where all the pods are connected to a single load balancer and service.

Resources

Project files are provided on github link. Clone the files using below link in terminal. $ git clone https://github.com/adeelbarki/capstone-clouddevops-docker-jenkins-aws-eks.git

Make sure to $ cd capstone-clouddevops-docker-jenkins-aws-eks in terminal to access all files.

Run the code

Create Stack:

• Create an EC2 instance with right permisions and user ID.

• Clone the project into the EC2 instance.

• Install all the required script. Script files can be found in folder 'scripts'. (Can be installed manualy too)

• Goto cfn folder and create stack using this command. aws cloudformation create-stack --stack-name eks-serv-node --region eu-central-1 --template-body file://amazon-eks-network.yml --parameters file://network-parameters.json --capabilities CAPABILITY_IAM

  Note: * Before running this command, AWS configure must be configured with a user Acces key and Id. * Single command will deploy network and server configuration on AWS. * Region can be selected according to your location. In this case eu-central-1 is selected. * AWS has limited instances in each region. In order to check your limit in specific region, goto EC2 > Limits in AWS console. This script is scaled to 4 instances and may not run if your EC2 limit is less than 4. In case of lower limit ask AWS help team to increase the limit.

• Configure EKS cluster with EC2 instance using this command aws eks update-kubeconfig --name nginxcluster

• Copy role arn from cloudformation stack outputs into aws-auth-cm.yml file.

• Configure aws-auth-cm.yml file in order to attach worker nodes to EKS cluster kubectl apply -f aws-auth-cm.yml

• Deploy initial code to kubernetes cluster and pods using blue deployment in Deployment kubectl apply -f blue-webapp-deploy.yml

• Initial Deployment is created. Now we just have to create the continuous pipepline

• Configure Jenkins and do not forget to install blue ocean agregator and pipeline-aws in jenkins plugins

• Configure global credentials for aws and docker hub.

• Create pipeline using github token; it will automatically detect the jenkinsfile.

• In the porcess we can see all the steps in the pipeline that inlcudes linting, build, deploy and blue/green strategies.

Testing

In EC2, just goto Load balancer and select specific load balancer. Copy provided DNS and open it in chrome.

DNS can also be obtained by typing kubectl get svc in command line.

Pods can be seen using command kubectl get pods and for nodes `kubectl get nodes

In order to test the pipeline, a new code is committed and it succefully update the website DNS in few minutes.

License

This Project is a part of Udacity DevOps Engineer Nano Degree Program.

Author

• Adeel Ahmed Khan (Adeel Barki)
  __ Full Stack Web Developer
  Front End Web Developer
  React Web Developer __