Project Overview

In this project, you will operationalize a Machine Learning Microservice API.

This project tests your ability to operationalize a Python flask app—in a provided file, app.py—that serves out predictions (inference) about housing prices through API calls.

Project Tasks

In this project you will:

• Test your project code using linting
• Deploy your containerized application using Docker and make a prediction
• Configure Kubernetes and create a Kubernetes cluster
• Deploy a container using Kubernetes and make a prediction

---

Project Steps

1. Setup the Environment

• Create a virtualenv, .devops, with command make setup it runs the command below:

python3 -m virtualenv --python=<path-to-Python3.7> .devops

• Activate the virtualenv, .devops, with command:

 source .devops/bin/activate

• Run make install to install the necessary dependencies in the requirements.txt & install hadolint using the Makefile commands:

pip install --upgrade pip &&\
  pip install -r requirements.txt
# Install hadolint
sudo wget -O /bin/hadolint https://github.com/hadolint/hadolint/releases/download/v1.16.3/hadolint-Linux-x86_64 &&\
  sudo chmod +x /bin/hadolint

• Install minikube using the commands:

curl -LO https://storage.googleapis.com/minikube/releases/latest/minikube-linux-amd64
sudo install minikube-linux-amd64 /usr/local/bin/minikube

• (Optional) You can create a symbolic link to minikube’s binary named ‘kubectl’:

ln -s $(which minikube) /usr/local/bin/kubectl

2. Containerize the Application using Docker

• The Dockerfile creates a working_directory /app/, copies the app.py to /app/, exposes port 80 and sets up the CMD to run app.py with a python interpreter when the container runs.

• Run make lint to lint the Dockerfileand app.py.

• Run run_docker.sh that builds a docker image tagged predictor, lists the docker images and runs predictorin a docker container.

• Run make_prediction.sh in a separate terminal to make the prediction.

• Run upload_docker.sh to tag and push the docker image created before to Docker Hub. You should change the variable dockerpath to your own account.

3. Configure & Create Kubernetes Cluster

• Start the minikube cluster using the command minikube start.

• Run run_kubernetes.sh that will deploy our docker image from docker hub to a minikube cluster, display the running pods and port forward port 8000 to 80.

• Run make_prediction.sh in a separate terminal to make the prediction.

• (Optional) You can delete the cluster using minikube delete or stop the cluster using minikube stop to conserve resources.