Datalab

CCL Home

Research

Papers
Projects
People
Jobs
REU

Software

Download
Manuals
Makeflow
Work Queue
Parrot
Chirp
SAND
AWE

Community

Highlights
Annual Meeting
Training
Getting Help
Mailing List
For Developers

Operations

Visual Status
Condor Pool
Hadoop Cluster
Biocompute
BXGrid
Condor Log Analyzer
Internal

Datalab

Datalab is an interactive platform for distributed scientific computing. Datalab lets you store sets of data files that are much larger than a single disk can handle, then define special functions to perform work on that data. Data processing is done in parallel on each storage node, so users reap the benefit of both distributed storage and processing in a way that outperforms conventional storage-only distributed computing systems.

More Info

Datalab User's Manual
Datalab Quick Start

Publications

DataLab Poster at HPDC 2008
DataLab: Transactional Data Parallel Computing on an Active Storage Cloud, Brandon Rich and Douglas Thain, IEEE/ACM High Performance Distributed Computing, June 2008. (two pages)
DataLab: Transactional Data-Parallel Computing on an Active Storage Cloud, Brandon Rich and Douglas Thain, Technical Report 2008-02, Department of Computer Science and Engineering, University of Notre Dame, March 2008. (ten pages)