The relevance of this study lies in the growing demand for high-load web applications, especially in e-commerce, social media, and streaming platforms, where performance, stability, and scalability are crucial. As user loads increase, traditional relational database encounter performance bottlenecks, highlighting the need for efficient caching solutions. Redis, a high-performance in-memory key-value store, is frequently used in such scenarios; however, the impact of different caching strategies on its performance remains understudied. The purpose of this article is to comprehensively evaluate the effectiveness of Redis as a caching tool for optimizing web application performance. The experimental design involved testing a web application backed by PostgreSQL under four conditions: no cache, Redis with cache-aside, Redis with write-through, and Redis Cluster. User loads of 1,000, 5,000, and 10,000 were simulated using Locust, and performance metrics were collected through Prometheus. Statistical analysis was performed using a t-test, and results were visualized with graphs and tables. The results show that Redis significantly decreases average response time (e.g., from 1146 ms to 323 ms in cache-aside mode), increases throughput (up to 226 requests/sec), and reduces the load on the main database. Cache-aside proved most effective for read-intensive workloads, while Redis Cluster offered better stability under high concurrency. The findings confirm Redis as a valuable component for high-load applications. Future research should explore Redis in distributed database settings, compare it to emerging tools like KeyDB, and examine its energy efficiency in cloud environments.
Liubomyr Kaptosv (Thu,) studied this question.