ENHANCING REAL-TIME DATA REPLICATION EFFICIENCY THROUGH OPTIMIZATION OF СHANGE DATA CAPTURE METHODS

Abstract

Relevance of the research is driven by the growing need for efficient real-time data replication in distributed and hybrid environments, where system scalability, consistency, and low latency are critical. Traditional change propagation mechanisms are increasingly inadequate under high-frequency workloads, highlighting the need for a re-evaluation and modernization of Change Data Capture (CDC) methods.The aim of the article is to provide a scientific rationale and develop approaches to improving the efficiency of real- time data replication through the optimization of Change Data Capture methods, taking into account the architectural and operational specifics of modern distributed computing environments.The research methodology is based on systems analysis of CDC implementations in cloud and hybrid infrastructures, architectural modeling of replication flows, comparative evaluation techniques, typological classification of functional characteristics, and a criterion-based approach to assessing efficiency in various data update scenarios.The research results include the identification of key functional features of CDC mechanisms, the classification of architectural replication models, and the substantiation of performance criteria such as latency, throughput, consistency, scalability, connectivity flexibility, and resource efficiency. Key technical constraints were identified in high-change-rate environments, particularly performance instability, access conflicts, and compatibility issues with traditional database systems and streaming platforms. It was demonstrated that event-driven architectures with asynchronous processing and adaptive buffering yield better performance when properly tuned.In the conclusions, the relevance of hybrid CDC models is substantiated, the dependency between replication efficiency and architectural processing models is confirmed, and common implementation barriers in conventional database and stream- processing ecosystems are outlined.The research perspectives include the development of intelligent CDC strategies, dynamic change flow orchestration, and the improvement of cross-platform interoperability in multi-cloud infrastructures.