Year: 2025
Type: article
Abstract: Cloud computing is essential for executing scientific workflows, offering scalable resources to process large volumes of data. However, efficient resource allocation remains a challenge due to fluctuations in demand and inadequate planning, which can increase makespan and thus energy consumption. In this context, infrastructure as a service is primarily used, allowing the rental of virtual machines with different characteristics. This study focuses on identifying the factors that influence the improvement of makespan for executing workflows from different disciplines and characteristics. The factors evaluated in this study are workfloF size, structure, and the number of virtual machines (VMs) required. The algorithms in this study consist of four heuristics and one metaheuristic, specifically a genetic algorithm (GA). Regarding the factors, the most relevant in CC is the number of VMs, where increasing the number of VMs reduces makespan to a certain extent, following the behavior described by the law of diminishing returns.
Authors: Elizabeth Del Angel Franco, Claudia Guadalupe Gómez-Santillán, Laura Cruz Reyes, Nelson Rangel Valdez, Gilberto Rivera Zarate
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
Funder: Tecnológico Nacional de México
OpenAlex ID: W4415089366
FWCI:
Cites:
Cited by:
Related to: