Operationalize a Machine Learning Microservice API - Kubernetes

Project Overview

Cloud DevOps Engineering

The premise and first instruction is - operationalizing a Machine Learning Microservice API with Kubernetes.

The entire project is based on a pre-trained, sklearn model that has been trained to predict housing prices in Boston according to several features, such as average rooms in a home and data about highway access, teacher-to-pupil ratios, and so on. You can read more about the data, which was initially taken from Kaggle, on the data source site.

This is a Python and Flask Implementation. The application app.py serves out predictions (inference) about housing prices through API calls. This project could be extended to any pre-trained machine learning model, such as those for image recognition and data labeling.

Project Tasks

The project goal is to operationalize this working, machine learning microservice using kubernetes, which is an open-source system for automating the management of containerized applications.

The project tasks are listed out below:

• Test project code using linting
• Complete a Dockerfile to containerize this application
• Deploy containerized application using Docker and make a prediction
• Improve the log statements in the source code for this application
• Configure Kubernetes and create a Kubernetes cluster
• Deploy a container using Kubernetes and make a prediction
• Upload a complete Github repo with CircleCI to indicate that code has been tested

You can find a detailed project rubric, here.

Note
The final implementation of the project will showcase abilities to operationalize production microservices.

---

Setup the Environment

Virtual Environments - It's recommended to leverage a virtual environment whenever using Python for projects. This keeps your dependencies for each project separate and organized. Instructions for setting up a virtual environment for your platform can be found in the python docs.

• Create a siloed virtual environment with Python 3.7 and activate it. You should have Python 3.7 available in your host/local machine. Check the Python path using which python3

python3 -m pip install --user virtualenv 
# use a command similar to this one to create environment:
python3 -m virtualenv --python=<path-to-python3.7> .devops
source .devops/bin/activate

Alternatively, you could setup the virtualenv via make setup. this is from a directive in Makefile.

• Run make install to install the necessary dependencies. This will install all relevant pip packages for the project.

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 via kubectl

---

Appendix

This holds information for some highlighted regular files & directory files in the root project repository.

appendix a - File Info:

• .circleci/config - Configuration file for CircleCI CI/CD tooling.
• model_data - Scikit Learn Dataset for the ML model.
• output_txt_files - Text output from CLI/shell commnds and/or scripts run against containers and clusters.
• app.py - The Python - Flask application that serves out API calls to the model.
• Dockerfile - Dockerfile with Python base image and commnds to run app in-container.
• Makefile - Makefile file for environment setup and linting in hadolint and pylint.
• requirements.txt - File for pip packages/dependencies.

appendix b - Helper Scripts:

make_prediction.sh run_docker.sh run_kubernetes.sh upload_docker.sh
These are bash scripts that you can use in place of multiple/chained shell commands.

resize.sh - A bash script to resize an AWS Cloud9 environment, if one opts to use a virtual machine for a flexible setup.

usage: run this in the terminal: bash resize.sh intended_cloud9_volume_size - size in Gigabytes.

running bash resize.sh uses 100GB by default.