In the realm of big data, where datasets of immense scale pose processing challenges, distributed processing platforms like open-source Apache Spark have emerged to address these issues. Spark’s internal configuration parameters exert varying impacts on execution times based on job characteristics, making manual optimization daunting. The core focus of this study lies in optimizing Spark’s internal configurations, with specific attention directed towards three types of workloads: Iterative-intensive, Memory-intensive, and CPU-intensive. Employing Grid Search, Random Search, and Evolutionary Optimization algorithms yields substantial execution time reductions: 23.24% with Grid Search, 19.71% with Random Search, and 23.06% with Evolutionary Optimization. Notably, Evolutionary Optimization achieves optimal configurations approximately 29% faster than Grid Search. While Random Search and Evolutionary Optimization share similar time requirements, Random Search’s execution time reduction for a given Spark workload is relatively lower. This research sheds light on algorithmic configuration tuning intricacies and its influence on Spark workload execution times, contributing to the exploration of optimizing big data processing platforms.