David P' Anderson

1
Designing Middleware for Volunteer Computing

• David P. Anderson
• Space Sciences Laboratory
• University of California Berkeley
• davea_at_ssl.berkeley.edu

2
Volunteer computing
Project start where area hosts GIMPS 1
994 math 10,000 distributed.net 1995 cry
ptography 100,000 SETI_at_home I 1999 UCB SETI
600,000 Folding_at_home 1999 Stanford biology
200,000 United Devices 2002 commercial biomedici
ne 200,000 CPDN 2003 Oxford climate
change 150,000 LHC_at_home 2004 CERN physics 6
0,000 Predictor_at_home 2004 Scripps biology 100
,000 WCG 2004 IBM biomedicine 200,000 Einst
ein_at_home 2005 LIGO astrophysics 200,000 SETI_at_h
ome II 2005 UCB SETI 850,000 Rosetta_at_home 2
005 U. Wash biology 100,000 SIMAP 2005 T.U.
Munich bioinformatics 10,000 Total of
BOINC-based projects 660,000 participants, 1,000
,000 hosts, 450 TeraFLOPS
3
Why volunteer computing?
home PCs
your computers
academic
business

• 1 billion PCs
• 55 privately owned
• most are on Internet
• If 100M participate
• gt 100 PetaFLOPs, 1 Exabyte (1018) storage
• Consumer products drive technology

4
What's different about volunteer computing?

• Must attract and retain volunteers
• Credit
• Community features
• Easy installation autonomic
• Volunteers are unreliable
• one solution redundant computing
• Heterogeneous, dynamic resource pool

5
Berkeley Open Infrastructure for Network
Computing (BOINC)

• Started in 2002 funded by NSF
• 2.75 FTEs lots of volunteers
• Open-source (LGPL)
• client 20K lines, C
• server 10K lines, C/Python
• web 10K lines, PHP
• http//boinc.berkeley.edu

6
Climate
SETI
physics
biomedical
projects
volunteers attach computers to
projects, allocate resources
volunteers
Joe
Jens
Alice
7
Client structure
servers
8
Server structure
Work creator
Transitioner
Scheduler
clients
Validator
Shared mem
MySQL
Assimilator
Feeder
File deleter
volunteers
DB purger
Web
1 server can handle 8-25 million tasks per day
9
Credit
10
Credit information flow
11
Goals of BOINC

• More projects
• Improve/simplify tools
• World Community Grid
• More participation
• Simplify everything
• GridRepublic
• Handle data-intensive apps better
• BitTorrent, use network topology
• Task graphs