Застосування гібридних алгоритмів сортування для оптимізації агрегації потокових даних моніторингу

Abstract

The study is dedicated to analyzing the efficiency of sorting algorithms in the context of processing large arrays of partially ordered data, typical for logging systems. A specific case of event log aggregation is examined, where records arrive with timestamps showing slight deviations from chronology. It is established that the use of adaptive hybrid methods allows for a reduction in processing time compared to classical quicksort algorithms by leveraging existing ordered subsequences.