Performance Comparison of Redis, Memcached, MySQL, and PostgreSQL: A Study on Key-Value and Relational Databases

This paper investigates and compares the performance of relational databases (MySQL and PostgreSQL) and key-value databases (Memcached and Redis) under various test loads. The study utilizes the Yahoo! Cloud Serving Benchmark to simulate diverse workloads and measure the behavior of these databases in different scenarios. The primary focus is on evaluating run time, throughput, and average latency metrics to understand how each database type handles varying thread levels and workload intensity. The outcomes of this research provide valuable insights into the scalability and efficiency aspects of relational databases. By conducting a comprehensive performance comparison, the study aims to assist database designers and developers in selecting the most suitable database option based on specific requirements. The findings contribute to informed decision-making regarding the choice between key-value and relational databases in various data storage scenarios.