A Comparative Performance Evaluation of SQLite, MySQL, and Firebase for Modern Application Development Using a Parallel Execution Approach

Abstract

This study conducted a performance evaluation of SQLite, MySQL, and Firebase (Firestore) databases using a parallel-execution approach. The rationale for the methodology was based on the necessity to maintain identical network and hardware conditions for the execution of queries across the various databases. By using this method, the performance metric concentrated exclusively on the execution time measured during different operations of Create, Read, Update, and Delete which was performed on both text and image data of varying payload sizes ranging from 50KB to 730KB. Results reveal several performance differences for different operations. For text-create and delete operations, Firebase recorded fastest average time (207 ms and 29.3 ms), while SQLite led in read operation (10.8 ms). The same trend was also recorded in working with image data of the same payload sizes. Firebase demonstrated the fastest create (32.3 ms) and delete (5.7 ms) times, while SQLite still attained the fasted read time (11,5 ms). Generally, the results of the experiment carried out on the Spring Boot framework reveals that application developed on top of firebase will perform faster than those developed on top of MySQL and SQLite in terms of data access time.