DisTInTo: Uma Ferramenta para Apoiar Experimentos Distribuídos em Larga Escala

• José Alexandre Macedo UFES
• Guilherme Tebaldi Meira UFES
• Rafael Fávaro UFES
• Guilherme Artem UFES
• Roberta Lima Gomes UFES
• Magnos Martinello UFES

Resumo

The scientific method dictates that experiments must be reproduced before they are considered valid. In computer science, reproduction requires open access to all code, scripts, and data used to produce the results. In particular, the validation of large scale distributed systems is a huge challenge to be performed in a real Internet-based testbed. In this work, we propose DiSTInTo as a tool designed to allow users to control their experiments at large-scale distributed testbeds. The tool supports the reproducibility of scientific experiments, since the setup instrumentation of machines until the crawling of the data. It provides features such as file transfer, experiments execution, real time monitoring and visualization of distributed machines. The time for experiments deployment, execution and crawling has been reduced in our P2P large scale networks tests.