Cloud DevOps ND - C4- Microservices at Scale using AWS & Kubernetes

This repository is associated with Cloud DevOps ND - Course 04 - Microservices at Scale using AWS & Kubernetes.

1. A summary of the project
2. Instructions on how to run the web app
3. A short explanation of the files in the repository

---

1. Summary

The parts of the project that have been completed by me are as follows:

1. Cloning this Repository
2. Creating and Activating the .devops Environment
3. Installing other Dependencies (Docker, Hadolint, Kubernetes, Minikube)
4. Completing the Project Tasks:
   1. Completing the Dockerfile
   2. Completing run_docker.sh
   3. Completing make_prediction.sh
   4. Improving the Logging of app.py
   5. Completing upload_docker.sh
   6. Configuring Kubernetes to Run Locally
   7. Completing run_kubernetes.sh
   8. CircleCI Integration
   9. Writing this README.md

---

2. Instructions

Firstly if you plan to upload the Docker image to your personal Docker Hub you have to add a docker_pw.txt file to the project-ml-microservice-kubernetes folder which contains only your Docker Hub password. Then change the vitaspm in upload_docker.sh to your own Docker Hub username.

To use the scripts in this repository, firstly change into the project folder using:
cd project-ml-microservice-kubernetes

Create a venv and install the dependencies using:
make setup
make install

Also install the additional dependencies: Docker, Hadolint, Kubernetes, Minikube

You can then:

• ./run_docker.sh to build a Docker container containing the web app
• ./upload_docker.sh to upload the Docker container to your Docker Hub account
• ./run_kubernetes.sh to run the Docker container on a Kubernetes cluster

Once the container with your app is running you can run ./make_prediction.sh to create a prediction using the app.

---

3. Files

Firstly the folders other than project-ml-microservice-kubernetes can be ignored since they are part of the repository but not related to the project. With the exception of .circleci which contains the config.yml which specifies the instructions of the CircleCI pipeline.

The project folder contains:

• app.py and model_data which are the necessary files for the prediction app to work
• output_txt_files which contains two logs of app tests using docker and kubernetes deployments
• Dockerfile which is used to build the Docker container to run the application
• Makefile which contains definitions of commands used by the circleci config and in the terminal
• requirements.txt which contains the libraries the app needs to work
• .sh files which can be used as explained in 2. Instructions