Title :
Implementing crossplatform distributed algorithms using standard web technologies
Author :
Martinez, Gina ; Val, Leonardo
Author_Institution :
Dept. de Inf. y Cienc. de la Comput., UCUDAL, Montevideo, Uruguay
Abstract :
This paper presents the development of a distributed computing framework based on standard Web technologies. The workload is specified as Javascript code, split into pieces served via HTTP and executed by web workers inside the clients´ Web browser. Initially developed to run genetic algorithms in a parallel fashion, our implementation has been generalized into a system powerful enough to run any kind of computationally intensive work. It allows running a distributed algorithm in a huge number of different devices. There are no special software requirements since every modern platform with a powerful CPU and a network connection has at least one Web browser available. The tests show tasks can be run in Web browsers in a secure and efficient manner. We show the great portability and performance achieved using a simple test bed problem.
Keywords :
Internet; Java; authoring languages; distributed algorithms; online front-ends; CPU; HTTP; Javascript code; Web browser; computationally intensive work; crossplatform distributed algorithms; distributed computing; network connection; standard Web technologies; Browsers; Computational modeling; Computers; Grid computing; Message systems; Servers; Software; Javascript; cross-platform; distributed algorithms; distributed computing; world wide web;
Conference_Titel :
Computing Conference (CLEI), 2014 XL Latin American
Conference_Location :
Montevideo
DOI :
10.1109/CLEI.2014.6965143