Project Overview

In this project, using a pre-trained, sklearn model that has been trained to predict housing prices in Boston, I operationalize a Python flask app using Docker and Kubernetes.

---

Files

• app.py - Flask microservice app containing pretrained model for predicting houseing prices in Boston.
• requirements.txt - Contains list of python modules the app needs installed to run properly.
• Makefile - utility file which defines a set of tasks to be run to build/test the application.
• Dockerfile - Dockerfile to build containerized environment for the application.
• run_docker.sh - Script to build and run the application inside a docker container.
• upload_docker.sh - Script to upload Docker image to Docker repository.
• run_kubernetes.sh - Script to pull application Docker image to deploy to Kubernetes cluster.
• make_prediction.sh - Script to send request to deployed application to make a prediction.
• .circleci/config.yml - CicleCI configuration file for config CI process.

Running app.py

1. Standalone: python app.py
2. Run app task: make [setup|install|test|lint|all]
3. Run in Docker: ./run_docker.sh
4. Upload Docker image: ./upload_docker.sh
5. Run in Kubernetes: ./run_kubernetes.sh