buklawz / operationalize-machine-learning-microservice-api

The project entails operationalizing a Python flask app—in a provided file, `app.py`—that serves out predictions (inference) about housing prices through API calls. The primary aim of the project is to containerize the python flask-app and deploy the container in Kubernetes cluster.

Project Overview

The project entails operationalizing a python flask app—in a provided file, app.py—that serves out predictions (inference) about housing prices through API calls. The primary aim of the project is to containerize the python flask-app and deploy the container in Kubernetes cluster.

Project Tasks

• Test project code using linting
• Complete a Dockerfile to containerize this application
• Deploy containerized application using Docker to Dockerhub
• Make a prediction by running the make_prediction.sh script
• Improve the log statements in the source code for the application
• Configure Kubernetes and create a Kubernetes cluster
• Pull the image from Dockerhub and deploy to local Kubernetes cluster
• Make a prediction by runiing the make_prediction.sh script
• Run circleci workflow

---

Setup the Environment

• Create a virtualenv and activate it

python3 -m venv <your_venv>
source <your_venv>/bin/activate

• Run make install to install the necessary dependencies

Running app.py

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

Kubernetes Steps

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

Files

• output_txt_files/docker_out.txt contains logs returned after running the app with Docker
• output_txt_files/kubernetes_out.txt containes logs and the prediction returned after running the app with Kubernetes(run_kubernetes.sh)
• run_docker.sh contains the steps to get Docker running the app locally
• run_kubernetes.sh contains the steps to get Kubernetes running the app locally
• upload_docker.sh contains the steps to upload the image to the Docker repository
• Screehshots of the different steps of executing the project are also included.