Estudio comparativo de mecanismos y algoritmos de balanceo de carga para un clúster Kubernetes de contenedores de bases de datos de solo lectura.
Loading...
Date
2024
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Universidad de Concepción
Abstract
This research focuses on performance optimization of Kubernetes clusters, with a special focus on read-only database containers. A cluster that emulates a real production environment was configured, implementing critical configurations for operation. Through the evaluation of multiple factors in various experimental scenarios, and using statistical methods such as ANOVA, Pareto analysis and interactions between factors, critical elements influencing performance were identified. The results revealed not only the most statistically significant factors, but also allowed, through a detailed analysis, to recognize the importance of those with minor but still significant effects, which are often underestimated because of their suboptimal contribution in the experimental design. This study concludes by underlining the relevance of a configuration tailored to the needs of the system, in addition to data-driven optimization to improve availability and efficiency under high-demand loads.
Description
Tesis presentada para optar al grado de Magíster en Ciencias de la Computación
Keywords
Contenedores, Algoritmos computacionales, Bases de datos