/intelligent-appraiser

Primary LanguageHTML

House Intelligent Appraiser Project

This project demonstrates how to operationalize an API using a Blue-Green CI/CD Jenkins pipeline deploying to a Docker containers in a Kubernetes pod. It implements the microservice API architecture to interact with a pre-trained machine learning model predicting the price of houses in the Boston, MA market.

The sklearn model has been trained to predict housing prices in Boston according to several features, such as average rooms in a home and data about highway access, teacher-to-pupil ratios, and so on. You can read more about the data, which was initially taken from Kaggle, on the data source site.

This project could be extended to any pre-trained machine learning model, such as those for image recognition and data labeling.

TL;DR

You will need an AWS account to run project:

  • Create an ec2 instance
  • install Jenkins
  • install Pipeline AWS: steps and Blue Ocean plugins for jenkins
  • install eksctl, docker, kubernetes, hadolint and pylint
  • Create a dockerhub account
  • Add the dockerhub credentials to Jenkins global credentials with id dockerhub
  • Create both AWS IAM User and Role with the necessary permissions to create an EKS cluster and nodegroup
  • Attach the role to the Jenkins server and add the IAM user credentials to Jenkins global credentials with id aws-bg-kube-pipeline
  • In ec2, select the Oregon region, us-west-2, and create a ppk called bg-pipeline
  • ssh into the server
  • run the following command eksctl create cluster --name prod --version 1.16 --region us-west-2 --nodegroup-name standard-workers --node-type t2.micro --nodes 1 --nodes-min 1 --nodes-max 1 --ssh-access --ssh-public-key bg-pipeline --managed
  • The command above takes about 15 minutes to run. You can navigate to CloudFormation to see all the resources it creates.
  • Fork this repository to your Github Account
  • Create a new pipeline in Blue Ocean and point it to this repository
  • Trigger a build to deploy the green container with its environment settings
  • Once the build is complete, run the following command from the terminal to get the URL of the Load Balancer kubectl get all
  • Navigate to that URL to see an SPA with a green background
  • Checkout the code
  • in Jenkinsfile, update the COLOR environment variable from green to blue, commit and push to Github
  • Trigger another run in Jenkins
  • Once the run is complete, refresh the your browser and this time the background will be blue

What You're Getting

├── CONTRIBUTING.md
├── README.md
├── Dockerfile
├── Jenkinsfile
├── app.py # the API for querying the model, it uses Flask 
├── blue_controller.json # Kubernetes ReplicationController settings for the blue environment
├── blue_index.html
├── blue_service.json # LoadBalancer settings for the blue environment
├── green_controller.json # Kubernetes ReplicationController settings for the green environment
├── green_index.html
├── green_service.json # LoadBalancer settings for the green environment
├── requirements.txt # Python dependencies to be installed in each container via pip
├── model_data
    ├── boston_housing_prediction.joblib
    └── housing.csv

Contributing

Do not hesitate to submit a pull request.

For details, check out CONTRIBUTING.md.