ДОСЛІДЖЕННЯ ВПЛИВУ НАЛАШТУВАНЬ КОНФІГУРАЦІЙНИХ ПАРАМЕТРІВ НА ПРОДУКТИВНІСТЬ РЕЖИМІВ РОЗГОРТАННЯ APACHE SPARK

The study is devoted to a comparative analysis of the application of different modes of using Apache Spark for the Word Count test dataset. An approach is proposed in which, based on the determination of the main configuration parameters of the Standalone and Hadoop Yarn modes and taking into account the limitations on the local deployment resource. A choice of configuration parameters and their changes in accordance with the default settings is proposed. The results are shown to substantiate its effectiveness based on the obtained quantitative estimates of the execution time.