|
ИСТИНА |
Войти в систему Регистрация |
Интеллектуальная Система Тематического Исследования НАукометрических данных |
||
Probabilistic Resources Allocation with Group Dependencies in Distributed Computingстатья
- Авторы: Toporkov Victor, Yemelyanov Dmitry, Bulkhak Artem
- Сборник: Lecture Notes in Computer Science
- Том: 14098
- Год издания: 2023
- Издательство: Springer Nature
- Местоположение издательства: Switzerland
- Первая страница: 151
- Последняя страница: 165
- DOI: 10.1007/978-3-031-41673-6_12
- Аннотация: In this work, we introduce and study a set of tree-based algorithms forresources allocation considering group dependencies between their parameters.Real world distributed and high-performance computing systems often operateunder conditions of the resources availability uncertainty caused by uncertaintiesof jobs execution, inaccuracies in runtime predictions and other global and localutilization events. In this way we can observe an availability over time functionfor each resource and use it as a scheduling parameter. As a single parallel jobusually occupies a set of resources, they shape groups with common probabilitiesof usage and release events. The novelty of the proposed approach is an efficientalgorithm considering groupings of resources by the common availability probability for the resources’ co-allocation. The proposed algorithm combines dynamicprogramming and greedy methods for the probability-based multiplicative knapsack problem with a tree-based branch and bounds approach. Simulation resultsand analysis are provided to compare different approaches, including greedy andbrute force solution.
- Добавил в систему: Топорков Виктор Васильевич