Citas bibligráficas
Esta es una referencia generada automáticamente. Modifíquela de ser necesario
Motta, A., (2024). State machine replication: exploring a path to build reliable services [Tesis, Universidad de Ingeniería y Tecnología]. https://hdl.handle.net/20.500.12815/380
Motta, A., State machine replication: exploring a path to build reliable services [Tesis]. PE: Universidad de Ingeniería y Tecnología; 2024. https://hdl.handle.net/20.500.12815/380
@misc{renati/231185,
title = "State machine replication: exploring a path to build reliable services",
author = "Motta Paz, Angel Alberto",
publisher = "Universidad de Ingeniería y Tecnología",
year = "2024"
}
Título: State machine replication: exploring a path to build reliable services
Autor(es): Motta Paz, Angel Alberto
Asesor(es): Bellido Angulo, Jesus Edwin
Palabras clave: Algoritmo de consenso; Método RABIA; Protocolos SMR; Sistemas distribuido; Tolerancia a fallas
Campo OCDE: https://purl.org/pe-repo/ocde/ford#1.02.02
Fecha de publicación: 2024
Institución: Universidad de Ingeniería y Tecnología
Resumen: Hoy en día disfrutamos de los beneficios de un mundo digital el cual nos brinda una creciente variedad de servicios. La alta disponibilidad de estos servicios es un aspecto importante y es posible gracias a la capacidad para tolerar fallas de los sistemas distribuidos. En este contexto los protocolos state machine replication (SMR) desempeñan un rol fundamental para implementar este tipo de servicios. Motivados por la importancia de los protocolos SMR, en este trabajo brindamos los fundamentos para comprender su funcionamiento y mostramos la alta complejidad inherente a estos protocolos así como sus principales causas y desafíos asociados. Producto de una revisión crítica de la literatura, identificamos una brecha de investigación de un nuevo método llamado RABIA, el cual tiene como objetivo simplificar el esfuerzo de implementación de un protocolo SMR utilizando la randomización en su algoritmo de consenso. Planteamos como propuesta profundizar el estudio de RABIA mediante un caso de estudio de implementación de un servicio de almacenamiento tipo llave-valor con capacidad de tolerancia a fallos y garantizando la consistencia de datos.
Currently we enjoy the benefits of a digital world that provides us with an increasing variety of services. The high availability of these services is an important aspect and is made possible by the fault tolerance of distributed systems. In this context, state machine replication protocols (SMR) play a fundamental role in implementing this type of services. Motivated by the importance of SMR protocols, in this paper we provide the fundamentals to understand their operation and show the high complexity inherent to these protocols as well as their main causes and associated challenges. As a result of a critical literature review, we identified a research gap for a new approach called RABIA, which aims to simplify the implementation effort of an SMR protocol by taking advantage of randomization in the consensus algorithm. We propose to deepen the study of RABIA by means of a case study which requires the implementation of a store layer based on keyvalue to offer a fault-tolerant service that guarantees data consistency.
Currently we enjoy the benefits of a digital world that provides us with an increasing variety of services. The high availability of these services is an important aspect and is made possible by the fault tolerance of distributed systems. In this context, state machine replication protocols (SMR) play a fundamental role in implementing this type of services. Motivated by the importance of SMR protocols, in this paper we provide the fundamentals to understand their operation and show the high complexity inherent to these protocols as well as their main causes and associated challenges. As a result of a critical literature review, we identified a research gap for a new approach called RABIA, which aims to simplify the implementation effort of an SMR protocol by taking advantage of randomization in the consensus algorithm. We propose to deepen the study of RABIA by means of a case study which requires the implementation of a store layer based on keyvalue to offer a fault-tolerant service that guarantees data consistency.
Enlace al repositorio: https://hdl.handle.net/20.500.12815/380
Disciplina académico-profesional: Ciencia de la Computación
Institución que otorga el grado o título: Universidad de Ingeniería y Tecnología. Ciencia de la Computación
Grado o título: Licenciado en Ciencia de la Computación
Jurado: Fiestas Iquira, José Antonio; Chambilla Aquino, Teofilo; Napa Bernuy, Angel Gerardo
Fecha de registro: 5-jul-2024
Este ítem está sujeto a una licencia Creative Commons Licencia Creative Commons