Database Performance Analysis in High-Concurrency Ticketing Systems

This research project implements and compares MySQL and MongoDB in a high-concurrency ticketing system environment, providing empirical evidence to support database selection decisions in similar applications.

Research Overview

The study evaluates how MySQL and MongoDB handle increasing concurrent loads (from 1 to 5,000 users) while managing ticket bookings, focusing on:

• Transactional behavior under high concurrency
• Impact of schema design on performance
• Management of nested data structures

Technical Implementation

The project uses Java to implement identical ticketing system functionality across both databases:

Core Technologies

• MySQL Community Server 8.0
• MongoDB Community Server 8.0.3
• OpenJDK 23.0.1
• Maven 3.9.5
• Hibernate (MySQL ORM)
• Morphia (MongoDB ODM)

System Architecture

The implementation follows a layered design:

• Data Access Layer: Database-specific implementations
• Service Layer: Business logic and transaction management
• Simulation Framework: Concurrent booking operations

Key Findings

The research revealed distinct performance characteristics:

• MySQL demonstrated superior consistency in transaction processing but with longer total processing times
• MongoDB showed better scalability under high load but with more variable performance patterns
• Schema modification impacts varied significantly between the two systems

Running the Tests

Detailed instructions for setting up and running the test scenarios are available in the docs directory.

Research Documentation

Full research findings, including detailed performance metrics and analysis, are available in the thesis document under the docs directory.

Author

Gabriel Alberto Avina Solares