Fair Ticketing System Web Application (TicketSlave)

Introduction

A web application that implements a fair ticketing system for concert events. The application allows users to buy tickets for events. The application also allows event organisers to create events and sell tickets for their events but with limited support.

Table of Contents

• Introduction
• Features
• Technologies
• Architecture
• Queuing Process
• Setup and Installation
• Usage
• Members

Features

• User Account Creation and Authentication: Secure user registration and login processes. Supports Google login
• Event Creation and Management: Intuitive tools for organisers to create and manage event details.
• Queueing System for Ticket Sales: Fair and transparent ticket allocation system to prevent scalping and bots.

Technologies

Backend

• Java Maven Spring Boot (Spring Security, Spring Data JPA, Spring Web)
• JUnit 5 (Mockito)
• Docker
• Apache Kafka
• Websockets

Frontend

• Svelte
• Tailwind CSS

Others

• AWS (S3, ECS, RDS, MSK, Cognito)
• Terraform
• Github Actions
• SonarCloud
• Stripe API

Architecture

Directory Structure

.
├── buffer-service
│   ├── buffer-backend
│   └── buffer-frontend
├── Deployment (organised using Terraform Modules)
│   ├── api_gateway
│   ├── cloudfront
│   ├── ecr
│   ├── ecs
│   ├── elb
│   ├── msk
│   ├── network
│   └── rds
├── feed-service
│   ├── feed-backend
│   └── feed-frontend
├── payment-service
│   ├── payment-backend
│   └── payment-frontend
├── purchase-service
│   ├── purchase-backend
│   └── purchase-frontend
└── queue-service
    ├── queue-backend
    └── queue-frontend

Architecture Diagram

Queuing Process

The queuing algorithm operates in two distinct phases:

1. Buffering and Shuffling: Initially, all incoming user requests are collected in a buffer. This collection phase is followed by a shuffling process where each user within the buffer is assigned a random queue number.
2. Sequential Processing: After shuffling, users are redirected to the queue service. Here, they are processed in a first-come, first-served manner, determined by their randomly assigned queue numbers. For users that join after the buffer is triggered, they will join from the back of the queue. This approach ensures fairness by randomising the order of service, rather than strictly processing requests in the order they were received.

Setup and Installation

Prerequisites

• Docker Desktop

Backend Setup

1. Ensure .env file is present with all the relevant environment variables. (not committed for security reasons)
2. From the root directory, start all services:

   docker-compose up

Frontend Setup

1. For each service, navigate to the frontend folder:

   cd [ServiceName]-service/[ServiceName]-frontend

2. Install dependencies and build the project:

   npm install && npm run build

3. After building, copy the index.html file generated in build directory into the corresponding backend service under resources/templates (e.g. index.html generated in feed-service/feed-frontend/build/_app should be copied to feed-service/feed-backend/src/main/resources/templates)

4. Upload the folder named _app to AWS S3. This will be the version served by the application.

5. Modify the source links in index.html to point to the correct S3 bucket and folder.

Usage

Upload newly built image to AWS ECR

1. Ensure you have the AWS CLI installed and configured with the correct credentials.
2. Ensure you have Docker installed and running.
3. From the root directory, run the following command:

   sh push_to_ecr.sh

Deploy AWS Infrastructure

1. Ensure you have the AWS CLI installed and configured with the correct credentials.
2. Ensure .env file is present with all the relevant environment variables. (not committed for security reasons)
3. From the root directory, run the following command:

   sh deploy.sh

Members

• Yue Zhen
• Mikhail
• Ryan Lim
• Teng Wee
• Linus
• Melanie