Publication: Diseño y desarrollo de una herramienta de simulación en Python para la composición de aplicaciones distribuidas de tiempo real basadas en servicios
Loading...
Identifiers
Publication date
2013-02
Defense date
2013-02-06
Authors
Advisors
Tutors
Journal Title
Journal ISSN
Volume Title
Publisher
Impact
Export
Abstract
En este proyecto n de carrera se aborda la composición de aplicaciones distribuidas de tiempo real
basadas en servicios. Para ello, se han implementado en lenguaje de programación Python dos algoritmos
de composición, uno exhaustivo y otro heurístico, para poder evaluar su comportamiento en un sistema
distribuido de tiempo real.
Estos algoritmos se han probado para aplicaciones sencillas, con pocas funcionalidades en serie, así
como aplicaciones más complejas en las que determinadas funcionalidades tengan que esperar a otras
para poder ejecutarse. Las aplicaciones, para ser compuestas, han de pasar una serie de condiciones y ser
plani cables en el sistema.
Para comprobar la plani cabilidad en el sistema se puede usar un algoritmo de plani cación exacto (con
tiempo de convergencia no acotado) o uno que calcule cotas al tiempo de respuesta (que da condiciones
su cientes pero no necesarias en la plani cabilidad). En concreto se ha elegido evaluar el comportamiento
del algoritmo heurístico con ambos algoritmos, así como la mejora que supone la introducción de un
control de acceso basado en cotas al tiempo de respuesta.
En cuanto al lenguaje de programación, se elige Python ya que es un lenguaje orientado a objetos y
que soporta diferentes plataformas, además de ser de fácil desarrollo y de tipado dinámico.
This nal project dissertation deals with the composition of real-time distributed applications based on services. On this purpose, two composition algorithms have been introduced in Python programming language, one exhaustive and the other one heuristic. It has been done in order to be able to evaluate its behaviour in a system distributed in real time. These algorithms have been tried in simple applications, with few functionalities in series, as well as in more complex applications in which these functionalities need to wait for other ones to be carried out. The applications shall go through a series of di erent conditions and be predictable in order to be compound. Either an exact schedulability algorithm (with not bounded convergence time) or another able to calculate the response time level (what gives us enough conditions but not necessary for the predictability) could be use to verify the system predictability. To be more speci c, it has been decided to evaluate the heuristic algorithm's behaviour with both schedulability algorithms, as well as the improvement what means the introduction of an access control system based on the bounded response-times. Regarding the programming language, Python has been selected not only because it supports multiple programming paradigms, including object oriented programming, and it is supported on di erent platforms but also because it uses dynamic typing, and application development is easy and fast.
This nal project dissertation deals with the composition of real-time distributed applications based on services. On this purpose, two composition algorithms have been introduced in Python programming language, one exhaustive and the other one heuristic. It has been done in order to be able to evaluate its behaviour in a system distributed in real time. These algorithms have been tried in simple applications, with few functionalities in series, as well as in more complex applications in which these functionalities need to wait for other ones to be carried out. The applications shall go through a series of di erent conditions and be predictable in order to be compound. Either an exact schedulability algorithm (with not bounded convergence time) or another able to calculate the response time level (what gives us enough conditions but not necessary for the predictability) could be use to verify the system predictability. To be more speci c, it has been decided to evaluate the heuristic algorithm's behaviour with both schedulability algorithms, as well as the improvement what means the introduction of an access control system based on the bounded response-times. Regarding the programming language, Python has been selected not only because it supports multiple programming paradigms, including object oriented programming, and it is supported on di erent platforms but also because it uses dynamic typing, and application development is easy and fast.
Description
Keywords
Python (Lenguaje de programación), Simulación, Sistemas en tiempo real, Sistemas distribuidos