Дослідження можливостей сучасних оркестраторів завдань для роботи із задачами обробки великих даних

Abstract

The purpose of this work is to overview and analyze main technologies that are used to orchestrate tasks in process of building Big Data processing pipelines, their advantages and disadvantages in order to propose the most efficient and flexible ways to organize them, monitor, measure and make them fault tolerant and idempotent.