This study examines the architectural principles for building fault-tolerant and high-performance real-time computing systems. The focus is on the use of design patterns such as Circuit Breaker, CQRS, and Saga Pattern, which contribute to the improvement of stability, performance, and data consistency in distributed systems. Performance optimization is achieved through asynchronous data processing, event-driven approaches, and the use of platforms such as Apache Kafka and Spark Streaming. Additionally, the importance of auto-scaling and effective caching is considered to ensure the stable operation of high-load systems.
Stoiev et al. (Mon,) studied this question.