Online Judge System API

• Author: J-Hoplin
• Team
  • J-Hoplin: Backend & Infrastructure
  • Oseungkwon: Frontend

Web Scraper

Use web scraper to fill problem and problem example datas. You need to run this scraper after you execute Online Judge System once.

• Repository

Contents

• 📦Diagram
• 📦Github Actions CI flow
• 📊Test Coverage
• 🧰Technical Stack
• ✅Run Application
• 🐳Run Application with Docker
• 🐰Run Application with Rabbit MQ worker locally
• 📄Run E2E Test
• 📄Run Unit Test
• 📝TODO

Diagram

You can either use AWS Environment or Docker envionment with this single repository

AWS Environment - AWS Elastic Beanstalk Worker

Docker environment - Rabbit MQ Worker

Github Actions CI flow

Frontend Repository

• Author: Oseungkwon
• Repository

Test Coverage

• E2E Test: 88.12%
• Unit & Integration Test: 79.13%

Technical Stack

• Language
  • TypeScript(Node.js v18 Runtime)
• Framework
  • Nest.js
• ORM
  • Prisma ORM
• Database(Persistence & Caching)
  • MySQL 8.0
  • Redis
  • AWS S3
• Issue Tracking
  • Sentry
• Proxy Server
  • Nginx
• Infrastructure
  • Docker & Docker-Compose
  • AWS Elastic Beanstalk(EC2 Instance)
    • Node.js Runtime x2 (Worker Server & Web Server)
    • Docker Runtime x1
    • AWS Worker Communication
  • AWS Auto Scaling Group
  • AWS SQS: For worker server
  • AWS S3: Build Versioning
• Test
  • Jest
  • Jest-Extended
• CI/CD
  • Github Actions
  • Code Pipeline & Code Build
• Alert
  • Discord

Run Application

1. Git clone repository

   git clone https://github.com/J-Hoplin/Online-Judge-System.git
   
   cd Online-Judge-System
   

2. Install dependencies

   yarn install
   

3. Run/Stop database with docker

   # Start
   yarn db:dev:up
   

   # Stop
   yarn db:dev:down
   

4. Sync prisma schema to database

   yarn db:push
   

5. Run application

   yarn dev
   

Run Application with docker

1. Build docker image

   docker build -t online-judge .
   

2. Run with docker enviornment

   yarn docker:up
   

3. Remove docker environment

   yarn docker:down
   

Run Application with Rabbit MQ locally

.env setting to use Docker Rabbit MQ worker

To use Rabbit MQ worker, you need to set both TYPE and QUEUE_TYPE as webserver and RMQ each. This is fundamental setting to use Rabbit MQ worker setting.

TYPE="worker"

...


QUEUE_TYPE="RMQ"
RMQ_URL="amqp://root:password@rmq:5672"
RMQ_WORKER_QUEUE_NAME="JUDGE_QUEUE"

...

To modify docker rabbit mq worker's .env file, modify .docker.worker.env. And if you want to modify docker webserver application, modify .docker.env.

How to build

1. Build docker image

   yarn docker:build
   

2. Run integrated application docker compose environment

   yarn docker:up
   

Run E2E Test

• Config: test/jest-e2e.json
• Mock Provider: test/mock.provider.ts

1. Run database

   yarn db:dev:up
   

2. Initialize test database

   yarn test:init
   

3. Run E2E Test

   yarn test:e2e
   

4. Run E2E Coverage Test

   yarn test:e2e:cov
   

Run Unit Test

• Config: test/jest-unit.json
• Mock Provider: test/mock.provider.ts

1. Run database

   yarn db:dev:up
   

2. Initialize test database

   yarn test:init
   

3. Run E2E Test

   yarn test:unit
   

4. Run E2E Coverage Test

   yarn test:unit:cov
   

TODO

• Apply Strategy Pattern to Asynchronous Worker
  • Use Nest.js Custom Provider
  • Rabbit MQ Strategy & AWS SQS Strategy
• Make Online Judge Server with Golang
  • Now using Judge0 based custom server