High Performance Cluster Architecture Using Free Software Tools
Leonardo Chuquiguanca, Edyson Malla, Freddy Ajila, Rene Guamán · Zenodo (CERN European Organization for Nuclear Research) · 2015
In this article the results of implementing a beowulf cluster type II architecture, with conventional hardware and free software tools is presented, allowed to convert a laboratory used for teaching programming systems, in a room with high performance, through a software: MPICH, Mosix, Blender, Hadoop and Ganglia Monitoring System. The method using for troubleshooting allowed; understanding the problem, and to design a solution strategy and achieve implementation of the cluster architecture with a performance like a supercomputer by optimizing the processing time: process simulation, pattern recognition, image and video rendering, and an analysis of Big Data, encryption codes, evaluation algorithms, etc. The tests allowed the processing of millions of operations per second reaching an efficiency of 85.2% of total capacity of cluster. This important achievement expedite also the academy investigation in others realms where required to process large volumes of information and get results in a short time.