Performance comparison and optimatization of cache based, document-based, column based and relational databases

In this study, cache-based, document-based, column-based and relational database architectures were selected for performance comparison. Real world query rows, hardware and software configurations were identical in virtiucal machine clusters; small, medium large and very large data volumes were read and complex query types were executed. Custom automation scripts completely automated the data collection, execution and reporting processes, ensuring consistency and repeatability of the tests. The collected metrics, execution times, latency and resource usage were visualized with tables and graphs, and then performance behaviors depending on the data volume and query structure were analyzed. The findings revealed that some architectures provided superior performance under certaion conditions, while others experienced bottlenecks as the sclae grew. Unique optimization suggestions were presented for the strengths and weaknesses of each architecture. This analysis provides guidance for performance-oriented decision-making processes in database selection. The study guides architectural preferences for performance requirements in different application scenarios.