Get-Youtube-Subscribers

A small Node.js REST API for managing YouTube-style subscriber records. It uses Express, MongoDB (via Mongoose), and automated tests, Docker, and CI/CD so you can run it locally, in a container, or deploy to Vercel after every push to main.

What this project covers

Topic	What you practice
Node.js & Express	HTTP routes, JSON APIs, environment variables (MONGO_URI)
MongoDB	Schemas, seeding data, integration tests against a real database
Docker	Packaging the app with a Dockerfile for consistent runs
CI/CD	GitHub Actions: test on every PR/push, deploy to Vercel on main

Tech stack

• Runtime: Node.js 20
• Framework: Express
• Database: MongoDB (Mongoose)
• Tests: Mocha, Chai, chai-http
• Deploy: Vercel (serverless Node via vercel.json)
• CI: GitHub Actions (.github/workflows/ci.yml)

Prerequisites

• Node.js 20.x
• MongoDB (local), or a cloud URI in .env
• For deployment: a Vercel project linked to this repo

Getting started (local)

1. Clone the repository:

git clone https://github.com/Arpit-Works/Get-Youtube-Subscribers.git
cd Get-Youtube-Subscribers

2. Install dependencies:

npm install

3. Configure environment (create .env in the project root):

MONGO_URI=mongodb://127.0.0.1:27017/subscribers
PORT=3000

4. Seed the database (MongoDB must be running):

node src/createDatabase.js

5. Start the server:

npm start

The API is available at http://localhost:3000.

API endpoints

Health check

• GET / — Returns API status and available routes.

Get all subscribers

• GET /subscribers — Array of subscriber documents.

Get subscriber names

• GET /subscribers/names — Subscribers with name and subscribedChannel only.

Get subscriber by ID

• GET /subscribers/:id
  • Valid ID: single subscriber object.
  • Invalid ID: 400 with { "message": "..." }.

Tests

npm test

In CI, tests run against a MongoDB service container on GitHub Actions: dependencies install with npm ci, the DB is seeded, then Mocha runs the suite.

Docker

The Dockerfile builds a minimal image: Node 20, npm ci, app code, and npm start on port 3000.

Build and run (set MONGO_URI to a host MongoDB or another container):

docker build -t get-youtube-subscribers .
docker run --rm -p 3000:3000 -e MONGO_URI=mongodb://host.docker.internal:27017/subscribers get-youtube-subscribers

.dockerignore keeps node_modules, .env files, and git metadata out of the image.

CI/CD (GitHub Actions + Vercel)

Workflow file: .github/workflows/ci.yml.

Pipeline flow

1. Test job (every push and pull request to main / master)

   • Spin up MongoDB
   • npm ci → seed DB → npm test

2. Deploy job (only on push to main, after tests pass)

   • vercel pull — production env and project settings
   • vercel build --prod — build artifacts in .vercel/output
   • vercel deploy --prebuilt --prod — upload prebuilt output to Vercel

GitHub secrets

Add these under Repository → Settings → Secrets and variables → Actions:

Secret	Description
VERCEL_TOKEN	Account token with access to your team (avoid project-only tokens for vercel pull)
VERCEL_ORG_ID	orgId from .vercel/project.json after vercel login and vercel link
VERCEL_PROJECT_ID	projectId from the same file (prj_…)

Do not commit .vercel/ or .env files; they are listed in .gitignore.

Lessons from setting up deploy

• vercel pull needs the right token scope — Project-scoped tokens often fail with “Could not retrieve Project Settings”; use a full account/team token when using the official pull → build → deploy --prebuilt flow.
• IDs are not names — VERCEL_ORG_ID is team_… or user_…, and VERCEL_PROJECT_ID is prj_…, not the project slug.
• Link locally once — Run vercel link to generate .vercel/project.json, then copy orgId and projectId into GitHub secrets.
• Separate test and deploy — PRs get tests only; production deploy runs only when main is green.

Project layout

├── index.js              # Vercel entry (routes to Express app)
├── vercel.json           # Vercel serverless config
├── src/
│   ├── index.js          # Local server entry
│   ├── app.js            # Express routes and MongoDB connection
│   ├── createDatabase.js # Seed script
│   └── models/           # Mongoose models
├── __tests__/            # API tests
├── scripts/run-tests.js
├── Dockerfile
└── .github/workflows/ci.yml

License

ISC