Learning Management System 📚

API only Rails app that handles online courses, its author and talents. The API includes CRUD operations for all these items. Also one important feature is that when deleting an author, the course gets transferred to another author. And if the new author was a talent of the transferred course then the course gets added to the list of authored_courses and removed from learnt_courses for the new author.

Also, I am not using any authentication library like Devise, we are just expecting the uuid for the users in requests.

Prerequisites

This app uses :-

• Ruby on Rails 6.0
• Ruby 2.6.3
• Postgres 13
• JSON API Resources from master
• Rspec for unit tests.

Installation

• git clone [email protected]:murtaza-bagwala/learning-management-system.git

• rvm use 2.6.3

• bundle i

• rename .env.example to .env and add the DATABASE_URL

• rails db:setup. This will create the DB, run the migration and seed the initial data from seeds.rb

• rails s

• rspec spec/. for testcases

or if you are familier with docker then just run these commands

• docker compose build
• docker compose up
• docker compose run app rake db:setup

note: docker-compose uses Dockerfile.local and there is another Dockefile which is created for fly.io deployment

Entity Relationship Diagram

An ERD Diagram, which mentions entities of the systems and their relationships. User can be an Author or a Talent, User has_many Authored Courses, Courses has_many Lessons, User has_many Learnt Courses through UserCourses and Course has_many Talents through UserCourses.

Postman Collection

To verify all the actions, postman collection is also added, you can find it in a root of a folder with a name Learning-Management-System.postman_collection.json, You can download the same and run on Postman collection runner. Default base_url is set as https://elearnio-rails-challenge.fly.dev this is where this application is currently deployed.

Current Application Infrastructure

Currently this application is deployed on fly.io on 512MB shared CPU with 1 GB Postgres DB storage but as we scale out we might need to move this to some big machine providers like AWS/GCP etc.

As I am a big proponent of extreme programming and CI/CD is one of the most important pillars of it, so I have setup the Github Actions workflows for the same.

CI

Before code gets merged to main branch, it runs 2 workflows.

1. Rubocop to findout any formatting issues.
2. Rspec unit tests to check if there are any failed testcases. I have written models and integration testcases.

CD

Now once CI passes successfully and code gets merged to main branch I have created another workflow .github/workflows/fly.yml which deploys the latest code to fly.io

Future Application Infrastructure

Now if scale out in future then this is how our deployment architecture would look like:-

• Terraform scripts to set the up the infrastructure.
• Once the code is merged Circle CI would build and pushes the image the ECR
• Trigger the EKS to pull the images from ECR and create the .containers