Hadoop for Roboticists - Archive ouverte HAL Access content directly
Reports (Research Report) Year : 2016

Hadoop for Roboticists

Abstract

The objective of this tutorial is to give an overview of Hadoop, its possibilities and its use in Robotics, as well as to provide explanations about the setup, design of mapreduce jobs and troubleshooting. Hadoop is a powerful software: while this tutorial is enough to get an overview and a better understanding of the framework and how to use it, we won’t be covering here all the possibilities that Hadoop offers. This tutorial is based on my experience while working on a Special Problem. I included all the information I have been looking for and everything I think can be useful when learning how to use Hadoop. Finally, while reading this tutorial, be aware that I am far from being a Hadoop expert and that I just started to learn about Hadoop. This is just a summary of the best of my knowledge at the time of writing. After presenting Hadoop and how to set up a cluster, we will mainly focus on developing jobs in C++. Although jobs are usually written in Java, there are applications for which we need to be able to use C++. One of such situations is for being able to use the ROS API in our jobs. This will allow us to analyze datasets stored as rosbags as one possible application.
Fichier principal
Vignette du fichier
CS8903_Deiss.pdf (319.77 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-01435882 , version 1 (20-01-2017)

Licence

Attribution

Identifiers

  • HAL Id : hal-01435882 , version 1

Cite

Olivier Deiss, Cedric Pradalier. Hadoop for Roboticists. [Research Report] UMI 2958 GeorgiaTech-CNRS. 2016. ⟨hal-01435882⟩
173 View
747 Download

Share

Gmail Facebook X LinkedIn More