Overview of approaches to scaling relational databases in development and adaptation of web applications

Abstract

The work explores various approaches to scaling relational databases in the development and adaptation of web applications. Relational databases, despite their reliability and stability, face challenges in handling large amounts of data and high concurrency. Key methods discussed include replication (master/slave and master/master), functional partitioning, horizontal data distribution (sharding), and denormalization. Each method offers specific advantages, such as improving performance, ensuring high availability, and providing fault tolerance, while also presenting certain disadvantages, including synchronization delays, data conflicts, or increased complexity. The study emphasizes the importance of selecting appropriate scaling techniques based on the unique requirements of a given application to ensure efficiency, scalability, and system stability under heavy loads.