Project Overview

The project is flask API application to provide predict housing prices in Boston which model has been trained based on the data source site.

---

Setup the Environment

# Setup virtualenv
make setup
source ~/.devops/activate

# Install dependencies
make install

# Execute Lint
make lint

Running app.py

1. Standalone: python app.py
2. Run in Docker: ./run_docker.sh
3. Run in Kubernetes: ./run_kubernetes.sh

Then, Send post request ./make_prediction.sh .

# Upload a docker image
docker login
./upload_docker.sh

Kubernetes Steps

• Setup and Configure Docker locally
• Setup and Configure Kubernetes locally
• Create Flask app in Container
• Run via kubectl

Directory

.
├── app.py             # Flask app
├── Dockerfile
├── Makefile
├── make_prediction.sh # a script to send POST data
├── model_data         # a directory for a model
├── output_txt_files   # outputs of logging
├── README.md
├── requirements.txt
├── run_docker.sh      # a script to run flask app as docker container
├── run_kubernetes.sh  # a script to run flask app as kubernetes cluster
└── upload_docker.sh   # a script to push docker images to a repository