Performance analysis of a small low cost beowulf cluster with mathematical modeling in a parallel computing environment

Abstract

Cluster computing has become the paradigm of choice for executing large scale science, engineering and commercial applications. This is due to their low cost, high performance, availability of off the shelf hardware components and freely accessible software tools that can be used for developing cluster applications. Commodity supercomputing clusters known as Beowulf clusters have become a low cost alternative to traditional supercomputers. Beowulf clusters combine inexpensive computers and specialized software to achieve supercomputing power. Beowulf clusters use commodity off-the-shelf hardware, and little information is available about their performance. This research includes the construction of a Beowuif cluster, research the available performance measurement schemes and measure the performance with increment of client nodes and changing memory at the nodes with mathematical modeling in a parallel computing environment. In addition, it is proposed to investigate the cluster technologies, system architecture, and the software tools for developing applications using low cost clusters in an environment for interacting with and steering large scale scientific computation running on clusters.