udacity-microservice

Summary:

This project builds and deploys a web-based House-price prediction microservice on Kubernetes.

Dependencies:

• python3
• minikube
• docker
• virtualbox

How to run the stand-alone webapp:

python3 app.py

How to package the webapp as container:

Create the virtual environment:

make setup

Upgrades pip and installs python libraries:

make install

lint python and Docker files:

make lint

shortcut for make install and make lint:

make all

How to deploy on docker:

Execute the script run_docker.sh. This will:

• Build an image tagged sklearn:v1
• Launch a docker container that exposes the microservice on port 8000

How to publish the new images:

Execute the upload_docker.sh This will:

• Authenticate and publish the image to hub.docker.com

How to deploy on kubernetes:

Execute run_kubernetes.sh This will: create deployment named sklearn create a service exposing the microservice externally

How to make a prediction:

Execute the script make_prediction.sh This will:

• Send a JSON formatted request to the microservice.