Publication: Generadores ScalaCheck para property-based testing de programas Spark y Spark Streaming
Loading...
Official URL
Full text at PDC
Publication Date
2016
Authors
Advisors (or tutors)
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
Citations
Exportar
Abstract
En los últimos años hemos sido testigos de la expansión del paradigma big data a una velocidad vertiginosa. Los cambios en este campo, nos permiten ampliar las áreas a tratar; lo que a su vez implica una mayor complejidad de los sistemas software asociados a estas tareas, como sucede en sistemas de monitorización o en el Internet de las Cosas (Internet of Things). Asimismo, la necesidad de implementar programas cada vez robustos y eficientes, es decir, que permitan el cómputo de datos a mayor velocidad y de los se obtengan información relevante, ahorrando costes y tiempo, ha propiciado la necesidad cada vez mayor de herramientas que permitan evaluar estos programas.
En este contexto, el presente proyecto se centra en extender la herramienta sscheck. Sscheck permite la generación de casos de prueba basados en propiedades de programas escritos en Spark y Spark Streaming. Estos lenguajes forman parte de un mismo marco de código abierto para la computación distribuida en clúster. Dado que las pruebas basadas en propiedades generan datos aleatorios, es difícil reproducir los problemas encontrados en una cierta sesion; por ello, la extensión se centrará en cargar y guardar casos de test en disco mediante el muestreo de datos desde colecciones mayores.
In recent years, we have been witness to the speedy expansion of big data. The changes in this field have led us to expand the treatable areas; which also implies more complex software systems associated to these tasks, as it happens in monitoring systems or in the Internet of Things. Likewise, the necessity of implementing reliable and efficient programs, that is, ones that allows us to compute data in a faster pace and to extract valuable information, saving money and time, has led to the increasing need of tools that permit us evaluate these programs. In this context, the present project centres its mission in extend the sscheck tool. Sscheck is able to generate test cases based in properties of programs wrote in Spark and Spark Streaming. These languages are part of the same open source cluster computing framework. Since the property-based tests generate random data, it’s complicated to reproduce the problems found in a given session; therefore, the extension will center in loading and saving test cases from disk through the sampling of data from bigger collections.
In recent years, we have been witness to the speedy expansion of big data. The changes in this field have led us to expand the treatable areas; which also implies more complex software systems associated to these tasks, as it happens in monitoring systems or in the Internet of Things. Likewise, the necessity of implementing reliable and efficient programs, that is, ones that allows us to compute data in a faster pace and to extract valuable information, saving money and time, has led to the increasing need of tools that permit us evaluate these programs. In this context, the present project centres its mission in extend the sscheck tool. Sscheck is able to generate test cases based in properties of programs wrote in Spark and Spark Streaming. These languages are part of the same open source cluster computing framework. Since the property-based tests generate random data, it’s complicated to reproduce the problems found in a given session; therefore, the extension will center in loading and saving test cases from disk through the sampling of data from bigger collections.
Description
Trabajo de Fin de Grado en Ingeniería Informática (Universidad Complutense, Facultad de Informática, curso 2015/2016)