Open Source Computer Cluster for Password Cracking

Marchand-Niño W.R.; Ciriaco-Soto G.; Ortiz-Morales E.F.; Olivera-Ruiz G.; Ortiz-Morales E.-A. · Aibi, Revista de Investigacion Administracion e Ingenierias (2025) · DOI: 10.15649/2346030x.4242

Resumen

This research focuses on the evaluation of efficiency in execution time for processes that demand significant computational capacity such as password cracking or password guessing from a hash, so the defined problem was, what is the level of efficiency of the open source computer cluster for password cracking processing for the purpose of validating password strength? For this purpose, a two-node computer cluster was used under a Linux operating system and OpenMPI, and the John the Ripper (JTR) tool was used for the password discovery process. The applied methodology includes four stages: selection of the hardware platform, selection of the software platform, cluster implementation and cluster testing, with two scenarios, testing with MD5 and SHA-1 hashes, and testing NT operating system hashes (Windows systems) and SHA-512 (Ubuntu 20.04 Linux operating system). The results of which indicate a decrease between 52% and 64% in the processing time to discover a password with the cluster at its maximum use of processors (16 in total) compared to the same procedure but with a single computer or node. Therefore, the level of efficiency found is positively significant.

Instituciones: Universidad Nacional Agraria de la Selva