from 27 September 2004 to 1 October 2004
Interlaken, Switzerland
Europe/Zurich timezone

HEP@HOME - A distributed computing system based on BOINC

oral presentation Track 5 - Distributed Computing Systems and Experiences Distributed Computing Systems and Experiences
30 Sep 2004, 14:00
Ballsaal (Interlaken, Switzerland)

Ballsaal

Interlaken, Switzerland

Description

Project SETI@HOME has proven to be one of the biggest successes of
distributed computing during the last years. With a quite simple
approach SETI manages to process huge amounts of data using a vast
amount of distributed computer power.

To extend the generic usage of these kinds of distributed computing
tools, BOINC (Berkeley Open Infrastructure for Network Computing) is
being developed. In this communication we propose a BOINC version
tailored to the specific requirements of the High Energy Physics (HEP)
community - the HEP@HOME

The HEP@HOME will be able to process large amounts of data
using virtually unlimited computing power, as BOINC does, and it should be
able to work according to HEP specifications.

One of the main applications of distributed computing is distributed data
analysis. In HEP the amounts of data to be analyzed have a large
order of magnitude. Therefore, one of the design principles of
this tool is to avoid data transfer - computation is done where data is
stored. This will allow scientists to run their analysis applications even
if they do not have a local copy of the data to be analyzed, taking
advantage of either very large farms of dedicated computers or
using their colleagues desktop PCs. This tool also satisfies other
important requirements in HEP, namely, security, fault-tolerance
and monitoring.

Primary authors

A. Amorim (FCUL Portugal / CERN) P. Andrade (University of Porto) J. Villate (FEUP Portugal)

Presentation Materials