Hadoop @ TACC

Welcome to Hadoop at TACC user group. This site is dedicated to the documents and guides to start up and run Apache Hadoop @ Texas Advanced Computing Center (TACC). Hadoop is an open source Java implementation of Google's MapReduce Distributed Computing Framework.

Dec. 18 2013 Notice: The Longhorn cluster will be decommissioned in early 2014. But don't panic. Another MapReduce system is coming soon to TACC. Please stay tuned for further updates.

Introductory slides

From High Performance Computing to Data-Intensive Computing: Analyzing Massive Datasets using Hadoop. Talk at TACC's 2011 Scientific Software Days event.

Hadoop on TACC Longhorn Cluster
- See various slides from CS395T / INF385T / LIN386M: Data-Intensive Computing for Text Analysis (Fall 2011)

Join the TACC-Hadoop email discussion list (Google group)

Longhorn Cluster Structure

Start-up guide

At a glance
- A simple recipe to run HADOOP at TACC
- Step-by-step Guide
- Create an user account at TACC
- Log in to Longhorn
- Install Hadoop at your TACC account
- Submit a job request
- VNC to TACC [Optional to view Job Tracker and Hadoop Admin]
- Advanced Users
- Longhorn User Guide

Get your hands dirty

Hello World! on Hadoop @ TACC
Upload your data to TACC computer
Word Count Example on Shakespeare's Othello
- PART I: Create Your Own Jar
- PART II: Run a Jar on Hadoop
- The super easy way to test your Hadoop installation and build jars, etc: Run wordcount with Taccdoop

Play-on guide

Trouble shooting

Backround: Hadoop

What is Hadoop? (Wikipedia)
Hadoop Wiki

Download Apache Hadoop
- How to setup at local computer for debugging and testing.
- Hadoop in Python - Dumbo
- How to set up Eclipse with hadoop (Courtesy of Dan Garrette)

Additional Information