PL/SQL Procedure Execution Latency in High Concurrency Banking Applications

Abstract

High-concurrency banking systems rely heavily on PLSQL procedures to enforce transactional
integrity and execute core financial operations in real time. However, as simultaneous user and
system-driven workloads increase, execution latency can escalate sharply, impacting throughput,
customer response times, and regulatory settlement deadlines. This study investigates the latency
behavior of PLSQL procedures under on-premise, private cloud, and public cloud deployments,
evaluating how procedure design, resource contention, and infrastructure configuration influence
performance. Experimental analysis shows that naïve row-by-row processing introduces early
contention amplification, whereas bulk-optimized set-based execution significantly delays latency
escalation. Additionally, cloud environments exhibit greater tail-latency variability, necessitating
adaptive routing strategies and workload-aware execution models. The findings highlight the
importance of concurrency-informed procedure design and environment-specific tuning to sustain
predictable performance in mission-critical banking applications.