Powered by OpenAIRE graph
Found an issue? Give us feedback

Capataz: a framework for distributing algorithms via the World Wide Web

Capataz: a framework for distributing algorithms via the World Wide Web

Abstract

In recent years, some scientists have embraced the distributed computing paradigm. As experiments and simulations demand ever more computing power, coordinating the efforts of many different processors is often the only reasonable resort. We developed an open-source distributed computing framework based on web technologies, and named it Capataz. Acting as an HTTP server, web browsers running on many different devices can connect to it to contribute in the execution of distributed algorithms written in Javascript. Capataz takes advantage of architectures with many cores using web workers. This paper presents an improvement in Capataz’ usability and why it was needed. In previous experiments the total time of distributed algorithms proved to be susceptible to changes in the execution time of the jobs. The system now adapts by bundling jobs together if they are too simple. The computational experiment to test the solution is a brute force estimation of pi. The benchmark results show that by bundling jobs, the overall perfomance is greatly increased. En los últimos años, algunos científicos han adoptado el paradigma de la computación distribuida. A medida que los experimentos y simulaciones demandan cada vez más poder computacional, coordinar varios procesadores es a menudo la única alternativa razonable. Hemos desarrollado una herramienta de computación distribuida de código abierto, basado en tecnologías web, que llamamos Capataz. Funcionando como un servidor HTTP, los navegadores que corren en diferentes dispositivos pueden conectarse para contribuir con la ejecución de algoritmos distribuidos escritos en Javascript. Capataz aprovecha las arquitecturas de múltiples núcleos mediante el uso de web workers. Este paper presenta una mejora a la usabilidad de Capataz y por qué es necesaria. En experimentos previos, el tiempo total de ejecución demostró ser susceptible a los cambios en el tiempo de ejecución de cada trabajo. El sistema ahora se adapta, empaquetando varios trabajos si éstos son demasiado simples. El experimento computacional para probar la solución es una estimación de fuerza bruta de pi. Los resultados prueban que al empaquetar los trabajos, el rendimiento se incrementa significativamente.

Keywords

distributed algorithms, world wide web, distributed computing, multiplataforma, Javascript, cross-platform, algoritmos distribuidos, computación distribuida

  • BIP!
    Impact byBIP!
    selected citations
    These citations are derived from selected sources.
    This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    0
    popularity
    This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
    Average
    influence
    This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    Average
    impulse
    This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
    Average
Powered by OpenAIRE graph
Found an issue? Give us feedback
selected citations
These citations are derived from selected sources.
This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Citations provided by BIP!
popularity
This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
BIP!Popularity provided by BIP!
influence
This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Influence provided by BIP!
impulse
This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
BIP!Impulse provided by BIP!
0
Average
Average
Average
Green