Benchmarking In-Memory vs. Disk-Based Databases: A Transaction Processing Performance Council (Tpc) Analysis

Abstract

Efficient online transactional processing (OLTP) applications hinge upon meticulous planning and optimization of database performance and throughput. The paramount objective for any database system deployed in this context is to minimize response times for application-specific structured query language (SQL) queries. This necessitates a strategic approach to address the challenges posed by concurrent user activity, which often leads to congestion and resource locking within the system. As highlighted by Kaspi and Venkatraman (2014) and TPC (2010), achieving optimal performance requires not only finely-tuned transactions but also robust features within relational database management systems (RDBMS) to manage these demands effectively. This study delves into the complexities of optimizing OLTP applications, emphasizing the critical importance of proactive planning and sophisticated database design to ensure seamless performance amidst concurrent user interactions. By synthesizing insights from seminal works by Faleiro and Abadi (2011) and leveraging established methodologies, the study offers actionable recommendations for enhancing the efficiency and scalability of OLTP-based systems. Key considerations include transaction tuning, resource allocation, and the integration of advanced features within RDBMS to mitigate congestion and streamline operations. By elucidating the multifaceted dynamics of OLTP optimization, this study contributes to a deeper understanding of the challenges and opportunities inherent in deploying transactional applications in online environments.