Desktop Grids for eScience

motherlamentationInternet και Εφαρμογές Web

7 Δεκ 2013 (πριν από 3 χρόνια και 10 μήνες)

179 εμφανίσεις

Desktop Grids
for eScience
Produced by the DEGISCO
project for the

International
Desktop
Grid

Federation
A Road Map
Detailed part – July 2011

2

IDGF/DEGISCO
International Desktop Grid federation
http://desktopgridfederation.org
Edited by
Ad Emmen
Leslie Versweyveld
Fabio Tumiatti
Oleg Sukhoroslov
Robert Lovas
Bernhard Schott
Graphics are produced by the projects.
version 2.2 2011-07-07
© 2011 DEGISCO Consortium: http://degisco.eu

DEGISCO is supported by the FP7 Capacities Programme under contract nr RI-261561.
Copyright (c) 2011. Members of DEGISCO consortium,
see http://degisco.eu/partners for details on the copyright holders.
You are permitted to copy and distribute verbatim copies of this document containing this copyright notice but modifying this document is
not allowed. You are permitted to copy this document in whole or in part into other documents if you attach the following reference to the
copied elements:
‘Copyright (c) 2011. Members of DEGISCO consortium - http://degisco.eu’.
The commercial use of any information contained in this document may require a license from the proprietor of that information.
The DEGISCO consortium members do not warrant that the information contained in the deliverable is capable of use, or that use of the
information is free from risk, and accept no liability for loss or damage suffered by any person and organisation using this information.

3

Preface
This document is produced by the DEGISCO
project for the International Desktop Grid Fed
-
eration.
Please note that there are some links in this
document pointing to the Desktop Grid Federa
-
tion portal, that requires you to be signed in first.
This is the companion document to a manage
-
ment document that serves as an introduction.

4

Table of Contents
Preface 3
Introduction 6
----------------------------------
Desktop Grids for eScience - a Road map -Details 7
1 Roadmaps 7
------------------------------------------------
1.1 How to start a Desktop Grid? 7
1.2 Road map for University (or research institute) to set up a local
----------------------------------------------------------------------
Desktop grid 8
----------------------------------
1.3 Road map for Volunteer Desktop Grids 11
1.4 Country politics: as part of a (Green) eScience
--------------------------------------------------------------------------
strategy 12
1.5 Companies to set up and participate as part of a "Corporate
-------------
Social Responsibility or Sustainable entrepreneurship 13
---------------------------------------------------------
2 Technical Documentation 17
-----------------------------------------------------
2.1 Desktop Grid Technologies 18
-----------------------------------------------------------------
2.1.1 Overview 18
-------------------------------------------------------------------
2.1.2 BOINC 18
---------------------------------------------
2.1.3 XtremWeb and XWHEP 19
-------------------------------------------------------------------
2.1.4 OurGrid 19
--------------------------------------------------------------------
2.1.5 Condor 20
-------------------------------------------
2.1.6 Commercial Technologies 20
---------------------------------------
2.2 Application Development and Porting 21
--------------------------------------------
2.2.1 Application Development 21
----------------------------
2.2.2 Porting Applications to Desktop Grids 22
-----------------------------------------------------
2.2.3 Application Portals 23
--
2.3 Integration of Desktop Grids and Other Computing Infrastructures 24
-----------------------------------------------------
2.3.1 EDGeS 3G Bridge 24
-----------------------------------------------------
2.3.2 Paths into Clouds 24
--------------------------
2.3.3 Collaboration with WEB2 technologies 25
----------------------------------------------------
3 Communication and marketing 26
--------------------------------------------
3.1Available communication channels 26
-----------------------------------------------------
3.1.1- BOINC channels 26
----------------------------------------------------------
3.1.2- IDGF channel 26
---------------------------------------
3.1.3- Country and local channels 27
-----------------------------
3.2 Available marketing and communication tools 28
----------------------------------------------------------------------
3.2.1 Portal 28
-------------------------------------------------------
3.2.2 Message boards 29
---------------------------------------------------------------------
3.2.3 Flyers 29
----------------------------------------------------------------------
3.2.4 Press 29
------------------------------------------------
3.2.5 Participation in events 29
------------------------------------------------------------
3.2.6 Social Media 29


5


---------------------------------------------------------------
3.2.7 Certified services 30
------------------------------------
3.3 Experiences in marketing desktop grids 30
--------------------------------------
3.3.1 CityGrid example: AlmereGrid 30
----------------------------------------
3.3.2 Public Grid: EDGeS@Home 31
---------------------------------
3.3.3 A national Desktop Grid: Ibercivis 32
-------------------------------------------
4 Legal and organisational information 33
---------------------------------------------------------------------
4.1 Privacy issues 34
----------------------------------------------------------------------------
4.2 Licensing 34
-------------------------------------------------------------------------------
4.3 Misuse 34
--------------------------------------------------------------------------
4.4 Warranties 34
-------------------------------------------------
4.5 Security based on certificates 35
-------
----------------------------------------------------------------------------
4.6 IPR 35
---------------------------------------------
4.7 Interoperation with Service Grids 35
----------------------------------------------------------------
5 Green Desktop Grids 36
-----------------------------------------
5.1 Relevance of Green Desktop Grids 36
-
5.1.1 Responsibility for real costs: Wise or waste use of energy 36
5.1.2 Selection of Green IT arguments: CO2 footprint and
--------------------------------------------------------
the energy mix 36
---------------------------------
5.1.3 € - a metric for Green IT success 36
--
5.1.4 The eleventh commandment: You shall not waste energy 37
------------------------------------------
5.2 Green Desktop Grid Methodologies 37
-------
5.2.1 The balance between complexity and Green features 37
--------------------------
5.2.2 Powerful organisational methodologies 37
--------------------------------------------------------------------
5.3 Implementation 39
-----------
5.3.1 The University of Westminster Local Desktop Grid 39
----------------
5.3.2 OurGrid the Brazil Desktop Grid saves energy 39
---
5.3.3 BOINC client configuration for time-of-Day energy tariffs 41
6 Glossary 43

6

Introduction
If you happened to be in search of a guide into
the emerging world of Desktop Grid Computing,
then you do not need to look any further for you
are about to discover the ultimate "Desktop
Grids for eScience" Companion.
Desktop Grids for eScience - a Road Map aims
to provide insight in the power of Desktop Grid
computing, what it can do for eScience, how it
can support scientists in the academic world,
researchers and engineers in industry, and how
citizens can actively be involved in this process.
As such, Desktop Grids offer a holistic approach
of providing huge amounts of computing power
for science. It is presented as a White Paper:
we try to present you correct, objective informa
-
tion.
The Road Map includes two parts: a manage
-
ment document, aimed at stakeholders in gov
-
ernment, academia and industry and a detailed
document, aimed at administrators and techni
-
cal operators in the field.
The management document provides a high
level and general overview and introduction of
Desktop Grid Computing for eScience. It offers
a catchy glimpse of what Desktop Grids have in
store for eScience by raising the appetite with a
few compelling success stories. A survey con
-
ducted to gauge the interest among the public
to donate idle computing time for eScience of
-
fers some remarkable results. The green as
-
pects of Desktop Grid computing are being
highlighted as well as the promising future de
-
velopments in Desktop Grids. And there is also
a short focus on how to bring Desktop Grid
computing to the general public with smart and
dedicated communication and marketing tools.
This detailed document is aimed at administra
-
tors and technical operators in the field.
This part of the road map is divided into five
sections.
The first section describes how to proceed to
set up a Desktop Grid. In fact, it provides sev
-
eral road maps for different stakeholders, i.e.
the academic world, volunteer Desktop Grids,
policy stakeholders at country level, and com
-
pany stakeholders to build out a corporate so
-
cial responsibility structure and a Desktop Grid
infrastructure that supports a sustainable entre
-
preneurship.
The second section is a technical description. It
offers an overview of the different Desktop Grid
technologies including a technology comparison
matrix. A Desktop Grid infrastructure also re
-
quires applications running on it. Therefore, this
chapter also elaborates on application devel
-
opment and application porting with a short de
-
scription of application portals. The second sec
-
tion concludes by explaining how Desktop Grids
can be integrated with other computing infra
-
structures.
The third section addresses communication and
marketing issues. Here, you can find an over
-
view of the communication channels and the
communication and marketing tools. This sec
-
tion also provides three case studies of Desktop
Grid marketing and promotion experiences.
The fourth section deals with the legal aspects
of Desktop Grid computing. It offers useful in
-
formation on organisational problems including
privacy, licensing, misuse, warranties, certifi
-
cates, IPR, and interoperation with Service
Grids.
The fifth section provides answers on how to
make and keep the Desktop Grid infrastructure
green. What options do we have to choose
from? Which are the challenges related to these
options?


7

1. Roadmaps
Desktop Grids can be used in different types of
organisations: in academia, in industry and in
groups of volunteer users or even at country
level.
In this section, you can find information about
the boundary conditions and policies to set up a
Desktop Grid.
The specific requirements for Desktop Grid
computing in the above mentioned environ
-
ments will be explained step by step.
1.1 How to start a Desktop Grid?
First of all, the boundary conditions and policies
must be clarified.
Deployment and operation of Desktop Grids
requires less efforts than other e-Infrastructures;
the barriers to entry of such systems are signifi
-
cantly lower.
Human resources: concerning the needed
skills, usually an system administration in part
time with average skills can operate a DG infra
-
structure. The labor cost depends on the coun
-
try and type of the organisation.
IT infrastructure: usually one server is the
expected hardware requirement, software is
usually for free due to the recommended open-
source software stack. The operational cost is
not negligible but much lower than the start-up
cost of hardware.
Operational and user policies: local policies
vary by organisation, please contact the local IT
department first
In case of a smaller volunteer project initiated
by a university with about 3,000 active hosts,
the majority of costs are for a 50%-time system
administrator/programmer. The total start-up
cost was about 3000 USD for a server with a
RAID storage server. The total monthly cost was
about 5000 USD/month.
In case of the largest BOINC based project;
SETI@HOME, the start-up cost is 43000 USD
and 12000 USD as monthly cost (including elec
-
tricity, power, personnel and other costs).
Please note that these costs do not cover the
operational and others expenditures related to
computers offered by volunteers or the other
shared computers in the campus/resource insti
-
tute or in the company.
Links
Derrick Kondo, Bahman Javadi, Paul Malecot, Franck
Cappello and David Anderson. Cost-Benefit Analysis of
Cloud Computing versus Desktop Grids. 18th Interna
-
tional Heterogeneity in Computing Workshop, May 25
2009, Rome.
http://mescal.imag.fr/membres/derrick.kondo/pubs/kondo_
hcw09.pdf
Distributed Computing in the Cloud - BOINC and Amazons
Elastic Compute Cloud EC2
http:/www.rechenkraft.net/wiki/index.php?title=Datei:Distri
buted_Computing_in_the_Cloud_-_BOINC_and_Amazons
_Elastic_Compute_Cloud_EC2_ENGLISCH.pdf
Steps/ sec
-
tions of the
roadmap
Select
goal
Study legal
require
-
ments 
Make
a plan
Select
user
group
Secure
financing 
Select
type
Select
technology 
Deploy/ Port

application
Implement
organisation 
Imple
-
ment
commu
-
nication
and sup
-
port
 General
 
4
 
 
 
2.2
2.1
2.2
2
3
 Universities/
Research
organisations
 1.2
Phase I 
 1.2 Phase I
(general)
5.1.2 Phase
II (local)
  1.2
Phase
I
 1.2
Phase II
 1.2
 Phase II
 3, 1.2
Phase V
(scale-up)
 1.2  Phase
III
1.2  Phase III
1.2  Phase III,
1.2 Phase V
 1.2 Phase
IV
Companies
 1.5
Phase II 
1.5 Phase I
(local)
1.5 Phase II
(general)
  1.5
Phase
II
1.5
Phase III 
1.5 Phase
III
-
1.5  Phase
IV
1.5  Phase IV
1.5 Phase IV
1.5 Phase
V

8

1.2 Road map for University (or re
-
search institute) to set up a local
Desktop grid
Phase I
Gather Desktop Grid related in
-
formation from IDGF sources
Phase I
Gather information and contact
the nearest IDGF representative
The most convenient way of starting and gath
-
ering information is reading this roadmap docu
-
ment and follow the links and references for
more detailed explanations. The nearest IDGF
representatives and experts can be easily
found in the Members section of the IDGF web
-
site.
Links
How-to-start section of International Desktop Grid Fed
-
eration Homepage
http://desktopgridfederation.org/how-to-start
Experts Section of International Desktop Grid Federa
-
tion Homepage
http://desktopgridfederation.org/experts
Join the IDGF
The benefits of IDGF membership can be taken
from the very beginning in the roadmap, that is
why, we recommend the interested ones to join
IDGF as soon as earliest.
Links
Membership application page on International Desktop
Grid Federation Homepage
http://desktopgridfederation.org/membership-application
Be self-trained or attend a half-day IDGF tu
-
torial
Numerous materials are available on-line for
self-training. However, we recommend to attend
at least a half-day tutorial given by IDGF experts
at various events or on-demand. For example, a
2-hour hands-on training is enough to perform
all steps of deployment of a typical Desktop Grid
with applications from scratch.
Studying the related best practices might be
very useful as well. For the most widespread
technologies LiveCDs, Virtual Images and demo
accounts are also provided for efficient self-
training.
All of these steps can help figure out and define
the preliminary goals and some plans.
Links
Tutorials on International Desktop Grid Federation
Homepage
http://desktopgridfederation.org/tutorials
Technical WIKI on International Desktop Grid Federation
Homepage
http://desktopgridfederation.org/technical-wiki
Phase II
Gather Desktop Grid related in
-
formation from local sources
Phase II
Collect local needs and local in
-
formation on opportunities
The primary motivation for having an own Desk
-
top Grid in a university or research institute is to
serve the local users with an easy-to-maintain,
green, and cost-efficient IT infrastructure. It is
very important to collect needs and application
ideas from the local community. Studying typical
applications scenarios and success stories can
help in this process. Also, it is important to
check the local conditions.
Links
Applications page on International Desktop Grid Federa
-
tion Homepage
http://desktopgridfederation.org/applications
Check local policies
Several different policies regulate the introduc
-
tion and operation of new IT solutions and serv
-
ices inside universities and research institutes.
The most widespread and usual requirements
are the following ones (among others):
The operation of the Desktop Grid should not
interfere with the teaching and research activi
-
ties in the laboratories, and should have negli
-
gible influence on the student experience or the
productivity of researchers.

9

The Desktop Grid system should work in-line
with the energy saving policy introduced by the
university or the research institute; e.g. in com
-
pletely idle mode the unattended PCs must be
shut down automatically.
Concerning safety regulations, the Desktop Grid
servers and clients must be handled similarly to
traditional servers if they are unattended (i.e.
fire alarm).
Introducing of the Desktop Grid must meet the
security requirements (firewall, protection of
user's data, etc.) as well, in some institutes the
peer-to-peer software solutions are handled
specially (CERN, MTA SZTAKI).
Links
Special configurations and policies in local desktop grid
environments on the International Desktop Grid Federa
-
tion Homepage (part of D2.1.1 deliverable)
http://desktopgridfederation.org/technical-wiki/-/wiki/Main/
Special+configurations+and+policies+in+local+desktop+gr
id+environments
Build team and allocate (start finding) re
-
sources
The team requires usually at least one expert in
system administration.
Phase III
Implementation
Phase III
Based on the requirements and capabilities,
after choosing the technology the infrastructure
must be designed together with allocation of
necessary hardware components. Since the
Desktop Grid systems try to harness and utilise
spare IT capacities, the extra hardware re
-
quirements are minimal.
Usually there are two steps during the deploy
-
ment of local Desktop Grids:
  - installation of server components
  - installation of clients
Links
Special configurations and policies in local desktop grid
environments' on the International Desktop Grid Federa
-
tion Homepage (part of D2.1.1 deliverable)
http://desktopgridfederation.org/technical-wiki/-/wiki/Main/
Special+configurations+and+policies+in+local+desktop+gr
id+environments
Introduce Desktop Grid policies
It can happen that some modifications are
needed in the local policies. On the other hand,
some typical policy documents, such as AUP
and SLA are provided that can be tailored ac
-
cording to the local circumstances.
Links
Acceptable use policy' on the International Desktop Grid
Federation Homepage (part of D2.1.1 deliverable)
http://desktopgridfederation.org/technical-wiki/-/wiki/Main/
Acceptable+use+policy
Deploy existing applications from the reposi
-
tories
Various tested applications are available from
the repositories in order to serve the local user
communities. If the required application cannot
be found in the repository; the application sup
-
port experts of IDGF are welcome the requests
tthrough the local teams or centrally
Links
Optional: deploy or use portals to access
resources
The easy access of computational resources
and applications are crucial from the exploita
-
tion point of view. Scientific gateways and por
-
tals can contribute to overcome such difficulties.
P-GRADE grid portal, WS-PGRADE portal, and
OurGrid portal are available for Desktop Grids.
Links
P-GRADE website
http://portal.p-grade.hu/
GUSE website
http://www.guse.hu
/
OurGrid public portal
http://portal.ourgrid.org/
Optional: use EADM for application porting
There are a growing number of applications
available in the repositories but there is an op
-
tion to port applications to Desktop Grids.
EDGeS Application Development Methodology
has been designed for this purpose and several
application developers have followed its steps
to get reliable and efficient applications.

10

During the application porting process the de
-
velopment, test, and validation infrastructure
can be used, provided by the IDGF core mem
-
bers or deployed locally using LiveCD or Virtual
Images.
Links
Report on best practices in application support (D3.3.1)
http://degisco.eu/documents/10522/38259/DEGISCO_D33
1_v19.pdf
Phase IV
Operation,training, communication
Phase IV
Train the users
The available User's Guides for the selected
technologies must be followed but the IDGF
experts are also available to give the necessary
trainings. A very frequent problem is that when
the users (e.g. biologists, chemists, or meteor
-
ologists) speak their own technical or scientific
jargon and the communication might be difficult.
The best trainers can speak the 'languages' of
both communities.
Links
Programming the Grid
http://desktopgridfederation.org/technical-wiki
Operate the infrastructure
The operation of infrastructure is lightweight
(similarly to the deployment) but the sysadmins
must care about some usual IT tasks, such as
monitoring
maintenance, and
local support.
Detailed guidelines are provided for these pur
-
poses.
Links
Report on best practices in infrastructure support (D2.4.1)
http://degisco.eu/documents/10522/38251/DEGISCO_D24
1_v17.pdf
Introduce Desktop Grid topics in curricula
(in case of universities)
Desktop Grids have been successfully inte
-
grated in the curricula of some technical higher
education institutes, summer schools, and they
are good candidates even for secondary
schools because of the relatively simple and
easy-to-understand approach towards the com
-
plex IT systems and solutions, and eScience
itself. The curricula can cover the two most im
-
portant topics:
applications
how to build Desktop Grids
Links
Joint EGEE and EDGeS Summer School on Grid Applica
-
tion Support
http://www.egee.hu/grid09/
Publish and disseminate results
The publication and communication of the
achieved scientific and other results can be pro
-
vided via various channels locally or by IDGF.
Links
5.3.1 Available communication channels
Please check this section of the document.
Phase V
Scale-up the infrastructure
Phase V
There are three options to involve more re
-
sources into the Desktop Grid.
Get worker or client nodes from other cam
-
pus(es) or department(s)
Colleagues from other campuses and depart
-
ment can be easily involved in the project.
Start volunteer project
This option is recommended after having an
operational and stable local desktop grid. The
volunteers must trust in the grid service opera
-
tor, i.e. the university or research institute, and
'going public' without solid background and ex
-
periences (at least in the local Desktop Grid
operations) is hazardous, and the volunteers'
trust might be lost.

11

That is why, it is highly recommended to vali
-
date the new applications (to be deployed in the
volunteer project) against the criteria defined in
the EADM methodologies.
On the other hand, according to the surveys the
application must be attractive from the volun
-
teer's point of view. They like supporting e.g.
medical applications but business applications
are usually not welcome.
The advertisement and public relations of volun
-
teer based Desktop Grids is also critical, some
technologies provide communication channels
towards the possible volunteers (BOINC).

Links
Technical Documentation
http://boinc.berkeley.edu/trac/wiki/ProjectMain
Integrate the infrastructure with service grids
There are two options in order to integrate the
created Desktop Grid infrastructure with service
grids.
The easiest way is to use the existing bridge
deployed by another partner and share jobs
between the gLite based Desktop Grid VO, and
the Desktop Grid itself.
Another option is to join another VO with the
Desktop Grid. Usually it requires more negotia
-
tion and a MoU between the VO and the Desk
-
top Grid operator, some examples are available.
In this case a 3G Bridge instance must be de
-
ployed and operated locally.

Links
3G Bridge' (part of D2.1.1 deliverable)
http://desktopgridfederation.org/technical-wiki/-/wiki/Main/3
G+Bridge

1.3 Road map for Volun
-
teer Desktop Grids
According to our experiences, it is recom
-
mended to get practised first by local Desktop
Grid technologies before launching a public vol
-
unteer project. The previous section describes
the roadmap for this, and only after the success
-
ful 'in-house' experiments it is recommended to
make it available for the wider audience.
Links
Volunteer Computing: The Ultimate Cloud By David P.
Anderson, ACM Crossroads, Spring 2010/Vol. 16, No. 3
http://portal.acm.org/citation.cfm?id=1734164&dl=ACM&c
oll=DL&CFID=7911886&CFTOKEN=41923379

12

1.4 Country politics: as
part of a (Green)
eScience strategy
The acceptance and penetration of using volun
-
teer and local desktop Grids as a daily routine in
the HPC and especially in the HTC related sci
-
entific communities are relatively low. Since the
user base of volunteer and local Desktop Grids
relies particularly on these communities, the
responsibility of funding agencies and higher
level decision makers is to two-fold:
catalyse the information exchange among the
key stakeholders, and
incorporate the volunteer and local desktop
Grids in the Green and eScience roadmaps/
strategies (and in their implementation) at
country-level more explicitly.
The rest of the section discusses these topics
based on related Roadmaps/Agendas.
In order to achieve the first goal, the invitation of
IDGF representatives for consultancies and
public information days organised by the fund
-
ing agencies and R&D related departments of
ministries may improve the communication.
Later the involvement of more representatives in
the related pan-European and worldwide sup
-
port organisation can be the next phase. These
steps are in-line with the
e-IRG roadmap
, sec
-
tion 3.4 'High Performance Computing' which
addresses 'to ensure all European expert com
-
munities can freely and efficiently share infor
-
mation, about both specific solutions and gen
-
eral best practices' concerning exascale com
-
puting know-how.
On the other hand, the recently published
DCI
Collaboration Roadmap
from the EGI, EDGI and
othere DCI projects, recognise the importance
of Desktop Grids as a part of Grid ecosystem,
and IDGF as key player in their support and
dissemination. It may move NGIs towards inte
-
gration of Desktop Grids with SGs that can help
implement green ICT nationally.
By following the
e-IRG roadmap
, section 3.8
'New User communities' which recommends the
membership countries, EC and the Grid opera
-
tors to provide mechanism for finding rapidly
available (financial) resources for the new users
communities. Together with the low barriers for
entry in Desktop Grids, it may accelerate the
adaptation of Desktop Grids (and therefore the
eInfrastructures themselves) by the user com
-
munities significantly faster.
The
Digital Agenda
, Section 2.7.1. 'ICT for envi
-
ronment' clearly states that 'The ICT sector
should lead the way by reporting its own envi
-
ronmental performance by adopting a common
measurement framework as a basis for setting
targets to reduce energy use and greenhouse
gas emissions of all processes involved in pro
-
duction, distribution, use and disposal of ICT
products and delivery of ICT services.' The
IDGF members are ready to cooperate in the
related Green computing actions and projects
nationally and internationally as well.
The Digital Agenda, Section 2.3 'Trust and se
-
curity' describes topics which are crucial for
Desktop Grids as well, the joint efforts in such
areas are also the common interests of the key
stakeholders in order to provide better security
mechanism for end-users and resource owners
(i.e. volunteers).
From a technical point of view, the efficient ex
-
ploitation of Desktop Grids depends on the
available internet bandwidth in a lot of cases.
The
Digital Agenda
, Section 2.4 'Fast and ultra
fast internet access' will let the volunteer Desk
-
top Grid operators allow to provide a more effi
-
cient computational platform for the scientists.
Joint actions with network service providers may
help attract more volunteers.
SIENA will support Europeʼs Distributed Com
-
puting Infrastructure (DCI) initiatives and the
European Commission in working towards the
delivery of a future e-Infrastructures roadmap
that will be aligned with the needs of European
and national initiatives. The mutual synchronisa
-
tion of these roadmaps and visions together
with the EDGI project will improve the sustain
-
ability of the defined recommendations.
Links
Digital Agenda for Europe
http://ec.europa.eu/information_society/digital-agenda/inde
x_en.htm
e-IRG Roadmap 2010
http://www.e-irg.eu/publications/roadmap.html
Distributed Computing Infrastructure (DCI) Collaborative
Roadmap
http://edgi-project.eu/documents/10515/50865/EDGID22v2.
pdf
SIENA Roadmap
http://www.sienainitiative.eu/StaticPage/Roadmap.aspx

13

1.5 Companies to set up
and participate as part
of a "Corporate Social
Responsibility or Sus
-
tainable entrepreneur
-
ship
Phase I
Check company policies
Phase I
Company polices restriction
One of the first things that should be done be
-
fore starting a Desktop Grid inside a company is
to contact the IT department and confirm that
the company doesn't have policies that prohibit
the use of Desktop Grids internally or any other
restriction about the Desktop Grid software li
-
cense. Some companies may have some re
-
strictions about using open source software or
software not registered internally.
Software license restriction
It is necessary to know the software that will be
executed/installed inside the Desktop Grid and
confirm with the IT department that the com
-
pany have the sufficient or adequate software
license to use it.
Links
Software license for running inside a Grid
http://www.elasticlm.de/fileadmin/SmartLM/Download/The
_Business_Side_of_Software_Licensing.pdf
Network connection
Some companies, for security reasons, recom
-
mend that the network connection between the
Desktop Grid machines should be a different
one from the main network used internally. Con
-
tact the IT department to know the security re
-
quirements of the company (firewall, protection
of user's data, etc.).
Phase II
Gather Grid related information
from IDGF sources
Phase II
Gather information and contact the nearest
IDGF representative
The most convenient way of starting and gath
-
ering information is reading this roadmap docu
-
ment and follow the links and references for
more detailed explanations. The nearest IDGF
representatives and experts can be easily
found in the Members section of the IDGF web
-
site.
Links
How-to-start section of International Desktop Grid Fed
-
eration Homepage
http://desktopgridfederation.org/how-to-start
Experts Section of International Desktop Grid Federation
Homepage
http://desktopgridfederation.org/experts
Join the IDGF
The benefits of IDGF membership can be taken
from the very beginning in the roadmap, that is
why we recommend the interested ones to join
IDGF as soon as earliest.
Links
Membership application page on International Desktop
Grid Federation Homepage
http://desktopgridfederation.org/membership-application
Be self-trained or attend a half-day IDGF tu
-
toria
l
Numerous materials are available on-line for
self-training. However, we recommend to attend
at least a half-day tutorial given by IDGF experts
at various events or on-demand. For example, a
2-hour hands-on training is enough to perform
all steps of deployment in a typical Desktop
Grids with applications from scratch.
Studying the related best practices might be
very useful as well. For the most widespread
technologies LiveCDs, Virtual Images and demo
accounts are also provided for efficient self-
training.
Links
Tutorials on International Desktop Grid Federation
Homepage
http://desktopgridfederation.org/tutorials
Technical WIKI on International Desktop Grid Federation
Homepage
http://desktopgridfederation.org/technical-wiki

14

Phase III
Gather Grid related information
from local sources
Phase III
Collect local needs and local information on
opportunities
The primary motivation for having an own Desk
-
top Grid in a company is to gain in computing
power, speed, and productivity. This leads to an
easy-to-maintain, green, and cost-efficient IT
infrastructure.
Some advantages for companies are:
No need to buy large and expensive servers
for running heavy/time costing applications.
Efficient on using idle resources. Many re
-
sources sit idle especially during off business
hours or are not using 100% of the computer
power consumption.
Modular environment: Don't have single
points of failure. If one of the desktops within
the Desktop Grid fails there are other resources
able to pick the load.
Scalability: Need more compute resources?
Just plug them in by installing Desktop Grid cli
-
ent on additional desktops. They can also be
removed easily.
Upgrading: Can be done on the fly without
scheduling downtime. Since there are so many
resources some can be taken offline while leav
-
ing enough for work to continue.
It is very important to collect needs and applica
-
tion ideas from the local community. Studying
typical applications scenarios and success sto
-
ries can help in this process. Also, it is important
to check the local conditions.
Links
Grid and Cloud Computing: A Business Perspective on
Technology and Applications
http://www.springer.com/business+%26+management/busi
ness+information+systems/book/978-3-642-05192-0
(chapter 3)
Build team and allocate (start finding) re
-
sources
The team requires usually at least one expert in
system administration.
Phase IV
Implementation
Phase IV
Based on the requirements and capabilities,
after choosing the technology the infrastructure
must be designed together with allocation of
necessary hardware components. Since the
Desktop Grid systems try to harness and utilise
spare IT capacities, the extra hardware re
-
quirements are minimal.
Usually there are two steps during the deploy
-
ment of local Desktop Grids:
installation of server components
installation of clients
Links
Special configurations and policies in local desktop grid
environments' on the International Desktop Grid Federa
-
tion Homepage (part of D2.1.1 deliverable)
http://desktopgridfederation.org/technical-wiki/-/wiki/Main/
Special+configurations+and+policies+in+local+desktop+gr
id+environments
Introduce Desktop Grid policies
It can happen that some modifications are
needed in the local policies. On the other hand,
some typical policy documents, such as AUP
and SLA are provided that can be tailored ac
-
cording to the local circumstances.
Links
Acceptable use policy on the International Desktop Grid
Federation Homepage (part of D2.1.1 deliverable)
http://desktopgridfederation.org/technical-wiki/-/wiki/Main/
Acceptable+use+policy
Deploy existing applications from the reposi
-
tories
Various tested applications are available from
the repositories in order to serve the local user
communities. If the required application cannot
be found in the repository; the application sup
-
port experts of IDGF welcome the requests
through the local teams or centrally.
Links
Membership application page on International Desktop
Grid Federation Homepage
http://desktopgridfederation.org/membership-application

15

Optional: deploy or use portals to access
resource
s
The easy access of computational resources
and applications are crucial from the exploita
-
tion point of view. Scientific gateways and por
-
tals can contribute to overcome such difficulties.
P-GRADE grid portal, WS-PGRADE portal, and
OurGrid portal are available for Desktop Grids.
Links
P-GRADE website
http://portal.p-grade.hu/
GUSE website
http://www.guse.hu
OurGrid public portal
http://portal.ourgrid.org/
Optional: use EADM for application porting
There are a growing number of applications
available of the repositories but there is an op
-
tion to port applications to Desktop Grids.
EDGeS Application Development Methodology
has been designed for this purpose and several
application developers have followed its steps
to get reliable and efficient applications.
During the application porting process the de
-
velopment, test, and validation infrastructure
can be used, provided by the IDGF core mem
-
bers or deployed locally using LiveCD or Virtual
Images.
Links
Report on best practices in application support (D3.3.1)
http://degisco.eu/documents/10522/38259/DEGISCO_D33
1_v19.pdf
Phase V
Operation, training, communication
Phase V
Train the users
The available User's Guides for the selected
technologies must be followed but the IDGF
experts are also available to give the necessary
trainings. A very frequent problem is that when
the users (e.g. biologists, chemists, or meteor
-
ologists) speak their own technical or scientific
jargon and the communication might be difficult.
The best trainers can speak the 'languages' of
both communities.
Links
Programming the Grid
http://desktopgridfederation.org/technical-wiki
Operate the infrastructure
The operation of infrastructure is lightweight
(similarly to the deployment) but the sysadmins
must care about some usual IT tasks, such as
monitoring
maintenance, and
local support.
Detailed guidelines are provided for these pur
-
poses.
Links
Report on best practices in infrastructure support (D2.4.1)
http://degisco.eu/documents/10522/38251/DEGISCO_D24
1_v17.pdf
Publish and disseminate results
The publication and communication of the
achieved scientific and other results can be via
various channels provided locally or by IDGF.
Links
Available communication channels
Please check this section of this document.
Phase VI
Participate as part of a Corpo
-
rate Social Responsibility
Phase VI
or Sustainable entrepreneurship
Corporate Social Responsibility
(CSR) can
make a significant contribution towards sustain
-
ability and competitiveness, both in Europe and
globally. The European Commission defines the
CSR as:
"A concept whereby companies integrate social
and environmental concerns in their business
operations and in their interaction with their
stakeholders on a voluntary basis."

16

Corporate Social Responsibility is part of the
Europe 2020 strategy for smart, sustainable and
inclusive growth. It can help to shape the kind of
competitiveness model that Europe wants.
The key issues of corporate social responsibility
vary from company to company. For example,
enterprises in the retail sector might have to
deal with the risk of poor labour standards in
their supply chain, while a mining company is
more concerned by the need to avoid infringing
the human rights of people living near its opera
-
tions.
Small and medium-sized enterprises (SMEs)
are the predominant form of enterprise in the
European Union. If Europe and its enterprises
are to reap the full benefits of CSR, it is vital to
make sure that SMEs are fully engaged and
that what they do is fully recognised.
Small businesses are typically not less respon
-
sible than large enterprises. They may not know
and use the term "CSR", but their close rela
-
tions with employees, the local community and
business partners often mean they have a natu
-
rally responsible approach to business.
CSR is more relevant than ever in the context
economic crisis. It can help to build (and rebuild)
trust in business, which is vital for the health of
Europe's social market economy. It can also
point the way to new forms of value of creation
based on addressing societal challenges, which
may represent a way out of the crisis.
Links
Sustainable and responsible business
http://ec.europa.eu/enterprise/policies/sustainable-busines
s/corporate-social-responsibility/index_en.htm
European business network for corporate social responsi
-
bility
http://www.csreurope.org/
Desktop Grid in the CSR
Basically at least two opportunities can be dis
-
tinguished which can improve the CSR of com
-
panies.
Some applications running on a Desktop Grid
have obviously direct social impact. Among oth
-
ers, these targeted applications can be listed in
this category:
City Population Dynamics and Sustainable
Growth (CPDynSG)
Drug discovery with Autodock
Social Simulations
BNB-GRID for Protein structure prediction
On the other hand, using Desktop Grid tech
-
nology has positive environment impact as well;
the company can significantly contribute to re
-
duce the ecological footprint by operating a
green Desktop Grid instead of other technolo
-
gies.
Links
World Community Grid
http://www.worldcommunitygrid.org
Success stories
Two success stories have been selected for
illustration purposes.
The first one is a global company, IBM, which
has been awarded by the Asia's Foremost CSR
Awards Programme. The award was given to
IBM Philippines (an ICPC country) in the cate
-
gory 'Poverty Alleviation' because of the related
efforts and achievements of BOINC-based IBM
World Community Grid.
The other example represents a local initiative;
Rabobank from the Netherlands shows the
AlmereGrid, the first citygrid in the world in its
demo room. Rabobank was a technology pro
-
vider for AlmereGrid in the BeINGRID project.
Links
Asia's Foremost CSR Awards Programme Awards Win
-
ners
http://www.chinaretailnews.com/2009/11/24/3168-asias-fore
most-csr-awards-programme-awards-winners-announced/

17

2. Technical Documenta
-
tion
The chapter with the technical documentation
starts with a description of the different existing
Desktop Grid technologies. They are compared
to each other in a technology comparison ma
-
trix. The next subpart addresses the applica
-
tions running on Desktop Grids: how to develop
applications and how to port them to the Desk
-
top Grid infrastructure. Here, you can also find
information on application portals. The chapter
ends by describing how you can integrate Desk
-
top Grids with other types of computing infra
-
structure, i.e. with the EDGeS 3G Bridge,
Clouds and WEB2 technologies.
BOINC
XtremWeb
OurGrid
Condor
Architecture
Client-Server
Client-Server
Peer-to-peer
Central Broker
Application Management
Centralized
Centralized
Decentralized
Decentralized
Resource Providers can
act as Resource Con
-
sumers
No
Yes
Yes
Yes
Task Distribution
Pull
Pull
Push
Push
Deployment/
Administration Complexity
Medium / Low on
client side
Low
Low
Medium
Application Development/
Porting Complexity
High / Medium (with
wrapper)
Low
Low
Medium
Support for Volunteer
DGs
Yes
Yes
Yes
No
Security Features
Code signing, Result
validation
Sandbox
Sandbox (Vir
-
tual Machine)
Authentication,
Authorization,
Encryption
Web Interface
Yes (Monitoring)
Yes (Monitor
-
ing)
Yes, via Our
-
Grid Portal
(Monitoring,
Job Submis
-
sion)
No
Distinguishing Features
Volunteer experience
Flexible ac
-
cess rights
model, Sup
-
port for X.509
certificates
Peer-to-peer
model,
Network-of-
Favors mecha
-
nism
Support for par
-
allel jobs (MPI,
PVM), work
-
flows (DAG)
Number of Deployments
~100
(~1M CPUs in big
projects)
~10
A few
~100
Programming Language
C/C++
Java
Java
C/C++
Documentation and
quickstart materials
Good
Good
Good
Very Good
Technology comparison matrix

18

2.1Desktop Grid Tech
-
nologies
2.1.1 Overview
Recommendations from DEGISCO Project de
-
liverable D2.4.1
In case of local desktop Grids where the pro
-
vided PCs are not from volunteers, the ap
-
proach of University of Westminster Local Desk
-
top Grid (WLDG) can be followed as best prac
-
tice that connects laboratory PCs of the Univer
-
sity of Westminster (London, UK) into a BOINC
based infrastructure.
In case of local desktop Grids where the re
-
source providers are particularly volunteers, the
OurGrid (peer-to-peer) or XtremWeb
HEP-
E
ex
-
amples can be followed.
In case of global desktop Grids, the approach of
SZTAKI Desktop Grid (for few applications) and
Ibercivis (for larger networks) can be followed
as best practice because of their success at
large scale as well.
2.1.2 BOINC
BOINC (Berkeley Open Infrastructure for Net
-
work Computing) is an open source platform for
Desktop Grid computing. It is being developed
at U.C. Berkeley Spaces Sciences Laboratory
by the group that developed and continues to
operate the SETI@home project. BOINC was
established in 2002 to develop general-purpose
middleware for volunteer computing, making it
easier and cheaper for scientists to use. The
first BOINC-based projects launched in 2004,
and today there are about 60 such projects, in a
wide range of scientific areas.
The BOINC software consists of two parts:
server software that is used to create volunteer
computing projects and client software. Anyone
can create a project. Each project operates its
own server and provides its own web site. Vol
-
unteers install and run client software on their
computers. The client software is available for
all major platforms, including Windows, Linux,
and Mac OS X.
A BOINC project is identified by a single master
URL, which is the home page of its web site and
also serves as a directory of scheduling servers.
A project can involve one or more applications,
and the set of applications can change over
time. BOINC accommodates a wide range of
applications; it provides flexible and scalable
mechanisms for distributing data, and its sched
-
uling algorithms intelligently match requirements
with resources. Existing applications in common
languages (C, C++, FORTRAN) can run as BO
-
INC applications with little or no modification.
Having installed the client programme, volun
-
teers can then attach it to any set of projects,
and for each project can assign a resource
share that determines how the computerʼs re
-
sources are divided among the projects. Attach
-
ing to a project allows it to run arbitrary execu
-
tables on oneʼs computer, and BOINC provides
only limited (account-based) sandboxing. So the
volunteer must assess the projectʼs authenticity,
its technical competence, and its scientific merit.
BOINC does accounting of credit, a numerical
measure of a volunteerʼs contribution to a pro
-
ject. It also provides a mechanism that exports
credit-related data for processing by credit sta
-
tistics sites operated by third parties. As part of
the accounting system, BOINC provides a
cross-project user identification mechanism
based on the volunteerʼs email address. This
mechanism allows leaderboard sites to display
user credit statistics summed over multiple pro
-
jects.
A level of indirection can be placed between
client and projects. Instead of being attached
directly to projects, the client can be attached to
a web service called an account manager. The
client periodically communicates with the ac
-
count manager, passing it account credentials
and receiving a list of projects to attach to. This
framework can be used for delegation of project
selection to some trusted committee, analogous
to mutual funds.
Links
BOINC Project Homepage
http://boinc.berkeley.edu/
User Documentation
http://boinc.berkeley.edu/wiki/Main_Page
Technical Documentation
http://boinc.berkeley.edu/trac/wiki/ProjectMain
Quick start: distributed computing in one hour or less
(Slides)
http://boinc.berkeley.edu/slides/boinc_intro.pdf
Quick start: distributed computing in one hour or less
(video)
http://video.google.com/videoplay?docid=-2165956560306
220048&hl=en
Installing a BOINC Grid
http://www.edges-grid.eu:8080/web/11219/2/wikipage/1197
8/InstallingBoincGrid

19

2.1.3 XtremWeb and XWHEP
XtremWeb is an open source software to build a
lightweight Desktop Grid by gathering the un
-
used resources of Desktop Computers (CPU,
storage, network). Its primary features permit
multi-users, multi-applications and cross-
domains deployments. XtremWeb was devel
-
oped by INRIA/IN2P3 as a research project to
explore scientific issues and applications of
Desktop Grid, Global Computing and Peer to
Peer distributed systems. XWHEP (XtremWeb-
HEP) is a successor of XtremWeb currently de
-
veloped at LAL (Laboratoire de lʼAccélérateur
Linéaire) belonging to IN2P3-CNRS. It is written
in Java and available under GPLv2 license.
The XWHEP/ XtremWeb architecture is com
-
posed of Servers, Workers and Clients. One or
a group of Servers is installed and maintained
by system administrators to host centralized
Services such as Scheduler and Result Collec
-
tor. Workers are installed by resource owners
on their PCs to propose their computing re
-
sources to be aggregated within an XWHEP/
XtremWeb infrastrucutre. Clients are installed
by resource users (scientists, for example) on
their PCs to interact with the infrastructure. The
Client permits users to manage the platform,
install applications and use distributed re
-
sources (submit jobs, retrieve results...). Jobs
submitted by Client are registered on the Server
and scheduled on Workers. Note that within this
architecture any Worker can be a Client.
XWHEP enforces greatly the security of
XtremWeb by introducing mechanisms aiming
to secure and confine distributed resources us
-
age. This mechanism is achieved by the imple
-
mentation of notions of user and access rights.
These new features permit to extend user ac
-
tions over the platform as well as to secure re
-
source usage and confine application deploy
-
ment.
Links
XtremWeb website
http://www.xtremweb.net/
XtremWeb-HEP website
http://www.xtremweb-hep.org/spip.php?lang=en
XWHEP User Guide
http://www.xtremweb-hep.org/lal/doc/xwhep-user-guide.pdf
XWHEP Admin Guide
http://www.xtremweb-hep.org/lal/doc/XWHEP-Admin-Instal
l.pdf
2.1.4 OurGrid
The OurGrid middleware enables the creation of
the so called peer-to-peer computational grids.
In the peer-to-peer grids enabled by OurGrid,
computing and storage resources originated
from spare resources (idle CPUs and free disk
space), are provided by a whole community of
grid participants, and are shared in such a way
as to make those who have contributed the
most to get the most out of the grid whenever
they need it. The software is written in Java,
allowing any resource capable of running a
Java Virtual Machine to be tapped on to the
grid. OurGrid is an open-source software dis
-
tributed under GPL license.
OurGrid provides a scalable and secure plat
-
form for execution of Bag-of-Tasks (BoT) appli
-
cations. BoT applications are those parallel ap
-
plications whose tasks are independent. These
include parameter sweep simulations, rendering
of images, optimization algorithms, and many
others.
OurGrid is based on a peer-to-peer architecture,
where each grid site corresponds to a peer in
the system. The main problem using a peer-to-
peer system is that the performance could be
compromised by freeriders: a peer that only
consumes resources, never contributing back to
the community. In order to incentive peers to
cooperate, and consequently discourage the
free riders, OurGrid uses a Network of Favours
mechanism: a favour is the allocation of a proc
-
essor to a peer that requests it, and the value of
that favor is the value of the work done for the
requesting peer. Each peer keeps a local record
of the total value of the favours it has given to
and received from any other peer. The rationale
is that each peer autonomously prioritize dona
-
tions to the peer to whom he owes most fa
-
vours, motivating cooperation.
OurGrid team also runs an open, free-to-join,
cooperative grid in which sites donate their idle
computational resources in exchange for ac
-
cessing other sites' idle resources when
needed. It is in production since December
2004. Any site can join the system by download
-
ing and installing the OurGrid software. No hu
-
man contacts or negotiation are needed for a
new site to join the system.
Links
OurGrid website
http://www.ourgrid.org/
User Documentation

20

http://www.ourgrid.org/index.php?option=com_content&vie
w=article&id=5&Itemid=2&lang=en
Developer Wiki
http://wiki.ourgrid.org/index.php/Main_Page
2.1.5 Condor
Condor is a high-throughput distributed batch
computing system developed by the Condor
Research Project at the University of Wisconsin-
Madison, USA. While providing functionality
similar to that of conventional batch systems,
Condor's novel architecture allows it to perform
well in environments where other batch systems
are weak: high-throughput computing and op
-
portunistic computing. The goal of a high-
throughput computing environment is to provide
large amounts of fault tolerant computational
power over prolonged periods of time by effec
-
tively utilizing all resources available to the net
-
work. The goal of opportunistic computing is the
ability to use resources whenever they are
available, without requiring one hundred percent
availability.
Condor can be used to manage a cluster of
dedicated compute nodes. In addition, unique
mechanisms enable Condor to effectively har
-
ness wasted CPU power from otherwise idle
desktop workstations. For instance, Condor can
be configured to only use desktop machines
where the keyboard and mouse are idle. Should
Condor detect that a machine is no longer
available (such as a key press detected), in
many circumstances Condor is able to trans
-
parently produce a checkpoint and migrate a job
to a different machine which would otherwise be
idle. Condor does not require a shared file sys
-
tem across machines - if no shared file system
is available, Condor can transfer the job's data
files on behalf of the user, or Condor may be
able to transparently redirect all the job's I/O
requests back to the submit machine. As a re
-
sult, Condor can be used to seamlessly com
-
bine all of an organisation's computational
power into one resource.
A Condor pool is comprised of a single machine
which serves as the central manager, and an
arbitrary number of other machines that have
joined the pool. Conceptually, the pool is a col
-
lection of resources (machines) and resource
requests (jobs). The role of Condor is to match
waiting requests with available resources. Every
part of Condor sends periodic updates to the
central manager, the centralized repository of
information about the state of the pool. Periodi
-
cally, the central manager assesses the current
state of the pool and tries to match pending re
-
quests with the appropriate resources.
The ClassAd mechanism in Condor provides an
extremely flexible and expressive framework for
matching resource requests (jobs) with resource
offers (machines). Jobs can easily state both job
requirements and job preferences. Likewise,
machines can specify requirements and prefer
-
ences about the jobs they are willing to run.
These requirements and preferences can be
described in powerful expressions, resulting in
Condor's adaptation to nearly any desired pol
-
icy.
Condor can be used to build Grid-style comput
-
ing environments that cross administrative
boundaries. Condor's "flocking" technology al
-
lows multiple Condor compute installations to
work together.
Links
Condor website
http://www.cs.wisc.edu/condor/
User and Administrator Documentation
http://www.cs.wisc.edu/condor/manual/
Publications on Condor
http://www.cs.wisc.edu/condor/publications.html
2.1.6 Commercial Technologies
Platform LSF Desktop
LSF Desktop is a Desktop Grid technology es
-
pecially built for company internal use. It is con
-
ceptually integrated into the Platform LSF prod
-
uct line and is professionally supported. The
LSF Desktop server is available for Linux and
appears as an LSF server with so many CPUs
(job slots) as there are desktops connecting to
the desktop server. The LSF Desktop client is
available for Microsoft Windows only. It uses
native Windows mechanisms for data protection
and job sandboxing. This is welcomed by Win
-
dows system administration as they can control
configuration and behaviour by policies in very
fine granularity. Most applications that run from
the Windows command-line will work with LSF
Desktop. The security mechanisms imple
-
mented are targeted to Enterprise use - inside
the secure company network only.
Inside the Enterprise, one could also use Plat
-
form LSF instead. LSF is available for all operat
-
ing systems – no restriction to Windows or Linux

21

– and runs perfectly on desktops. Different from
LSF Desktop, the priority for the workload can
be chosen within the boundaries of the selected
OS. Depending on the importance of the work
-
load and the commitment of the desktop owner,
higher priorities can be chosen compared to
“screen-saver” – enabling quality of service for
workload. MPI and workflows are natively sup
-
ported.
Platform computing
http://www.platform.com/workload-management/high-perfo
rmance-computing
DataSynapse / TIBCO GridServer
Desktop Grid computing provider DataSynapse
was acquired by Tibco in 2009 for US$28 mil
-
lion. Until then the company's installed base
was about 100 worldwide organisations running
DataSynapse application management software
across approximately 1 million CPUs making it,
probably, the most successful desktop grid mid
-
dleware after BOINC.
Nowadays DataSynapse software is provided
as a TIBCO GridServer which is one of the mar
-
ket's leading grid computing infrastructure plat
-
forms. With more than one million deployed
CPUs spread across 1,000 global installations,
TIBCO GridServer is the backbone of enterprise
grids deployed by businesses operating in the
world's most demanding industries.
TIBCO Grid server
http://prod-staging.tibco.com/products/soa/grid-computing/
gridserver/default.jsp
Parabon Frontier
http://www.parabon.com/
ProActive Parallel Suite
http://www.activeeon.com/products
Digipede Network
http://www.digipede.net/
Red Hat MRG Grid
http://www.redhat.com/mrg/grid/
Univa Grid MP
http://www.univa.com/products/grid-mp.php
2.2. Application Devel
-
opment and Porting
2.2.1 Application Development
Applications suitable for running on a desktop
grid should meet the following requirements:
Only master/worker or parameter sweep
parallelisation (bag-of-tasks applications)
No shared data storage
No MPI or internal communication between
worker nodes
Nodes can use the results of other nodes,
but only through the server
Typically long running jobs with small or
medium-sized (max. 100 MB per slave) inputs
and outputs
Applications running on desktop grids are gen
-
erally composed of two distributed parts: server-
side (master part) and client-side (worker part).
The server-side part of application is responsi
-
ble for generating tasks to be sent to desktop
machines, validating results of task execution
and processing these results into some form
presented to the application user.
The client-side part runs on desktop machines
and is responsible for executing tasks. It must
be able to run as a background process without
interactive input or GUI. The heterogeneity of
desktop grid nodes implies that the client-side
part of the application should be available for all
popular platforms.
The desktop grid technology provides a mid
-
dleware layer between these two parts of appli
-
cation which transparently distributes applica
-
tion code and tasks among grid machines,
monitors task execution, handles failures, col
-
lects task results and passes them to the appli
-
cation.
Desktop grid technologies provide generic
Command-Line (CLI) or Application Program
-
ming Interfaces (API) for application develop
-
ment, deployment and management.
For example, BOINC provides both server-side
(work unit generation, result validation and as
-
similation) and client-side (application execution

22

control, checkpointing and resource usage ac
-
counting) interfaces. Since there are many leg
-
acy applications, that have either no source
code available or would require too much effort
to port, BOINC also provides a configurable
wrapper which lets you run any executable (or
sequence of executables) with no modification.
GenWrapper is an advanced generic solution
for wrapping and executing an arbitrary set of
legacy applications for BOINC.
Both XtremWeb/XWHEP and OurGrid support
execution of arbitrary binary files which simpli
-
fies application development.
In case of XtremWeb/XWHEP, an application
developer should first register an application
executable by uploading it on XtremWeb Server.
It is possible to register multiple application bi
-
naries for different target platforms. Any user
can register its own applications, but the appli
-
cation access rights depend on the user creden
-
tials. Administrator rights are needed to register
a public application accessible by all users. Af
-
ter the application is registered a user can sub
-
mit new jobs by providing an application identi
-
fier, command-line arguments and references to
input data. Bag-of-tasks applications are sup
-
ported by managing groups of jobs.
OurGrid deals with BoT applications, i.e. appli
-
cations comprised of independent tasks. An ap
-
plication is described by a user in the form of job
description file. A job consists of multiple tasks.
A task is formed by initial, remote, and final sub
-
tasks, which are executed sequentially. Sub
-
tasks are external commands invoked by Our
-
Grid; consequently, any programme, written in
any language, can be a subtask. The initial and
final subtasks are executed locally, on the user
machine. The initial subtask is meant to set up
the taskʼs environment by, for instance, transfer
-
ring the input data to the grid machine selected
to run the task. The final subtask is typically
used to collect the taskʼs results back to the
user machine. The remote subtask runs on a
grid machine and performs the computation per
se. In addition to its subtasks, a task definition
also includes the grid machine requirements.
OurGrid allow users to write subtasks without
knowing details about the grid machines on
which they will be run, such as how their file
systems are organized. The abstractions stor
-
age and playpen represent storage spaces in
the unknown grid machine. File transfer com
-
mands are available to send and retrieve files in
the storage spaces.
Condor supports both execution of unmodified
binary files as well as programmes relinked with
Condor library which provides automatic check
-
pointing and remote system calls. Controlling
the details of a job submission is a submit de
-
scription file. The file contains information about
the job such as what executable to run, the files
to use for keyboard and screen data, the plat
-
form type required to run the programme, and
where to send e-mail when the job completes.
You can also tell Condor how many times to run
a programme; it is simple to run the same pro
-
gram multiple times with multiple data sets.
The Distributed Computing API (DC-API) was
created by MTA SZTAKI to allow easy imple
-
mentation and deployment of distributed appli
-
cations on multiple grid environments. In order
to accommodate the needs of very different grid
environments, the DC-API supports only a re
-
stricted master-worker programming model. The
application developed using DC-API can be run
in BOINC and Condor environments. The API
comes prepacked in a Debian package, avail
-
able freely from the official website indicated
below.
Links
BOINC Application Developement
http://boinc.berkeley.edu/trac/wiki/ProjectMain
BOINC Application Developement
http://boinc.berkeley.edu/trac/wiki/AppIntro
GenWrapper
http://genwrapper.sourceforge.net/
XWHEP User Guide
http://www.xtremweb-hep.org/lal/doc/xwhep-user-guide.pdf
Running a Job in OurGrid
http://www.ourgrid.org/index.php?option=com_content&vie
w=article&id=61&Itemid=54&lang=en
Road-map for Running Jobs in Condor
http://www.cs.wisc.edu/condor/manual/v7.5/2_4Road_map
_Running.html
Application Development (SZTAKI Desktop Grid)
http://desktopgrid.hu/index.php?page=22
DCI-API
http://www.desktopgrid.hu/index.php?page=34

2.2.2 Porting Applications to Desktop Grids
Although many applications have been ported
to the grid or developed specifically to be run on
various distributed computing resources, these
efforts were in many cases ad-hoc and did not
follow any particular methodology. In most of the
Grid projects middleware development was at

23

the focus, and application development served
in many cases demonstrative purposes only.
Even in examples of production level application
development efforts, these applications were
very often poorly documented which made fur
-
ther improvement cumbersome. In order to
overcome these difficulties and to identify a
process and a series of documentation tem
-
plates, the EDGeS project has developed the
EDGeS Application Development Methodology
(EADM). This methodology was successfully
validated and utilized in the EDGeS project
when porting 22 target applications, including
popular ones such as AutoDock, to the EDGeS
Grid infrastructure.
EADM identifies the following well defined
stages that have a suggested logical order:
1.
Analysis of current application
2.
Requirements analysis
3.
Systems design
4.
Detailed design
5.
Implementation
6.
Testing
7.
Validation
8.
Deployment
9.
User support, maintenance & feedback
The overall development process in most cases
is non-linear allowing revisiting and revising the
results of previous phases at any point. Each
stage is described by its objectives, participants,
outcome and set of questions to be answered.
EADM also defines how the recommended
software tools (DC-API, GenWrapper, WS-
PGRADE portal, EDGeS Application Reposi
-
tory) can aid the process of application devel
-
opment and execution in Desktop Grids.
Links
List of applications ported to Desktop Grid infrastructures
http://desktopgridfederation.org/applications
Application Porting Introduction
http://desktopgridfederation.org/documents/10508/22303/I
DGF-Application-porting-introduction.pdf
EDGeS Application Development Methodology
http://portal.desktopgridfederation.org/documents/10508/8
0409/EADM.ppt
2.2.3 Application Portals
The P-GRADE Grid Portal is a web based,
feature-rich environment for the development,
execution and monitoring of workflows and
workflow based parameter studies on various
grid platforms. P-GRADE Portal hides low-level
grid access mechanisms by high-level graphical
interfaces, making even non grid expert users
capable of defining and executing distributed
applications on multi-institutional computing in
-
frastructures. Workflows and workflow based
parameter studies defined in the P-GRADE Por
-
tal are portable between grid platforms without
learning new systems or re-engineering pro
-
gramme code. Technology neutral interfaces
and concepts of the P-GRADE Portal help users
cope with the large variety of grid solutions.
More than that, any P-GRADE Portal installation
can access multiple grids simultaneously, which
enables the easy distribution of complex appli
-
cations on several platforms.
gUSE (grid User Support Environment) is an
evolution of P-GRADE which is implemented as
a set of Web services that bind together in flexi
-
ble ways on demand to deliver user services in
Grid and/or Web services environments. User
interfaces for gUSE services are provided by
the WS-PGRADE Web application. WS-
PGRADE is a Web portal hosted in a standard
portal framework. WS-PGRADE uses the client
APIs of gUSE services to turn user requests into
sequences of gUSE specific Web service calls.
WS-PGRADE hides the communication proto
-
cols and sequences behind JSR168 compliant
portlets. End users can access WS-PGRADE
via Web browsers. A graph editor component
can be downloaded from WS-PGRADE via the
browser to the user machine. The editor can be
used to define the static skeleton of workflows,
while the HTML pages of WS-PGRADE provide
interfaces to add content to graphs, to generate
complete Grid/Web service applications.
OurGrid developers provide a public portal
which enables users to run jobs in OurGrid
Community opportunistic grid without installing
and configuring OurGrid software on their ma
-
chines. The Portal also provides a data storage
service where users can place application and
input files. Output files generated by jobs can
also be retrieved using this user-friendly service.

24

Links
P-GRADE website
http://portal.p-grade.hu/
GUSE website
http://www.guse.hu
OurGrid public portal
http://portal.ourgrid.org/
Security Issues
There are several security risks associated with
Desktop Grids. The most important one is exe
-
cution of malicious code on donor machines.
This issue is taken in account by all major Desk
-
top Grid technologies. The best practices de
-
pend on the amount of trust between donors
and application developers/users.
Private and local grids (eg. Condor-based)
where such trust is high usually employ only
standard security mechanisms such as authen
-
tication, authorization and encryption. Central
-
ized volunteer computing projects (eg. BOINC-
based) assume that volunteers trust project ap
-
plications and therefore use digitally signed ex
-
ecutables in order to prevent distribution of ma
-
licious code.
Peer-to-peer Desktop Grids that allow any user
to run its own application (eg. based on OurGrid
or XtremWeb) have low amount of trust be
-
tween participants and therefore should use
more advanced security mechanisms. The most
widely used of such mechanisms is sandboxing
by running untrusted executable inside a re
-
stricted environment or a virtual machine. This
mechanism introduces additional management
and performance overhead but ensures that the
host system won't be damaged by executable
from Desktop Grid.
Another security measure that can prevent un
-
trusted applications from running on Desktop
Grid is the use of Application Repositories that
provide an authoritative source for trusted and
validated applications.
Links
Security Issues
http://boinc.berkeley.edu/trac/wiki/SecurityIssu
es
2.3 Integration of Desk
-
top Grids and Other
Computing Infrastruc
-
tures
2.3.1 EDGeS 3G Bridge
The Generic Grid-Grid Bridge (3G Bridge) is a
software component used within the EDGeS
project that provides the core component of the
Service Grid - Desktop Grid interoperability so
-
lution. This technology enables one to combine
these two types of grid systems in order to es
-
tablish a seamless and vast grid resource pool.
In particular, the software is being used to
bridge EGEE service grid with the BOINC and
XtremWeb desktop grids.
Links
Policies and setup operations for Grid operators
http://desktopgridfederation.org/technical-wiki/-/wiki/Main/
Policies+and+setup+operations+for+Grid+operators
General information on the EDGeS Bridge
http://edges-grid.eu/web/edges/57
2.3.2 Paths into Clouds
Cloud computing is en vogue. Clouds are impor
-
tant infrastructures or building blocks of infra
-
structures. Hence it is important to start investi
-
gating the impact that cloud computing will have
on Desktop Grid computing.
The integration of and collaboration between
Grids and Clouds is investigated in several
European projects. The Siena initiative is work
-
ing together with them to create a Road map on
a future e-Infrastructure Road map that includes
Grids and Clouds.
One step, explored by for instance LHC@Home
and XtremWeb
HEP-
E
is to package applications
in standardized Virtual machine (VM) and let the
Desktop Grid infrastructure handle VM's instead
of applications. This opens a whole range of
questions on security, stability and validation.
Another approach is being implemented by the
EDGI project. The idea is to implement small
Clouds of always available clusters that can be
used to be able to guarantee a reasonable SLA
level for users.

25

Links
Integrated ARC, Desktop Grid, and Eucalyptus bridge.
http://edgi-project.eu/documents/10515/50865/EDGID61.pdf
Siena initiative
http://www.sienainitiative.eu/
2.3.3 Collaboration with Web 2.0 technolo
-
gies
The main aim of the Hungarian Web2Grid pro
-
ject is to support Web 2.0 communities in dy
-
namically building the necessary infrastructure
that is needed for processing the high-end tasks
of their community, the same way as the content
is generated as joint effort of the members
Web2Grid technology enables the community to
build the required infrastructure as a joint com
-
munity effort exploiting the volunteer computing
resources of the community members.
In order to achieve this goal Web2Grid exploits
the BOINC and SZDG technologies and other
earlier initiatives focusing on the integration of
desktop grid and Web 2.0 technologies. On the
one hand, the Web 2.0 technologies assist to
advertise desktop grid goals and attract compu
-
tational resources for desktop grid communities.
On the other hand, the developed new
Web2Grid platform that includes desktop grid
technology can offer back-end infrastructure for
several compute-intensive Web 2.0 related
problems .
The combination of the Web 2.0 and desktop
grid technologies provides advantages for both
sides, they can mutually benefit from each
otherʼs approach. In such combination Web 2.0
can extend its capabilities further from the
community contents towards to shared services
with the help of grid technologies.
The Web2Grid project developed a demo Face
-
book application for image watermarking, which
utilizes a desktop grid system for running wa
-
termarking jobs and an accounting system han
-
dling micro-payments. The underlying desktop
grid system is SZDG that is represented by a
web service behind which there is a workflow
enactor system called gUSE, a bridge and a
volunteer desktop grid system based on BOINC.
Links
WWeb2Grid Project Description on MTA SZTAKI home
page
http://www.sztaki.hu/search/projects/project_information/?
uid=00274
WWeb2Grid project website (in Hungarian)
http://web2grid.econet.hu/
Wikipedia entry
http://en.wikipedia.org/wiki/WWeb2Grid

26

3. Communication and
marketing
Communicating and marketing Desktop Grids
involves addressing several audiences:
Potential volunteers
Scientific users
System administrators
Funding and other agencies
Press
The EDGeS project conducted a survey about
what people expect from a Desktop Grid. This
can be used as a basis for your communication
plan.
In this chapter we describe three aspects of
communicating about Desktop Grids:
Available communication channels
This part describes the organisations that al
-
ready do communication about Desktop Grids.
Available marketing and communication tools
This part describes the tools, such as web por
-
tal, flyers, etc.
Experiences in marketing desktop grids
This part describes a number of examples on
how communication tools are used by some
typical Desktop Grids.
Links
Acceptance of Desktop Grid computing amongst SME's
and the General Public; Ad Emmen, Leslie Versweyveld,
Stichting AlmereGrid, Almere, The Netherlands
http://www.edges-grid.eu:8080/c/document_library/get_file
?p_l_id=11065&folderId=11075&name=DLFE-1027.pdf
3.1

Available communi
-
cation channels
3.1.1 BOINC channels
BOINC is the only Desktop Grid middleware that
not only provides software, but an ecosystem of
users at all levels around it.
On the message boards of BOINC also sugges
-
tions for communicating and advertising projects
can be found.
There is regularly a BOINC workshop organ
-
ised.
If you set up a volunteer desktop grid, it can be
useful to get it listed on one of the BOINC statis
-
tics websites.
Links
BOINC developer website
http://boinc.berkeley.edu/trac/wiki/WikiStart
BOINC message boards
http://ibercivis.eshttp://boinc.berkeley.edu/dev/
BOINC workshops
http://boinc.berkeley.edu/trac/wiki/BoincEvents
BOINCstats
http://boincstats.com/
3.1.2 IDGF channe
l
When you become member of the International
Desktop Grid federation, you can benefit from
the marketing and communication channels of
that organisation.
First, you apply for IDGF membership by filling
out an application form. Once accepted by the
IDGF community and having paid the member
-
ship fee, you can enjoy several services that are
provided at the IDGF web portal.
The International Desktop Grid federation brings
together experts from all areas related to Desk
-
top Grid computing. The community also in
-
cludes people who are familiar with integrating
desktops into scientific infrastructures. As a
member, you are part of this network and can
benefit from their expertise.
The IDGF portal hosts a library with flyers,
presentations given at conferences and work
-

27

shops, press material, and technical documen
-
tation. There is also an up-to-date news archive.
In the tutorial section, you can download the
educational material from past tutorials and
subscribe to future tutorials. Material from past
workshops is available in the events section.
Here you can download presentations given at
previous workshops and conferences and regis
-
ter for upcoming conferences and workshops.
In the infrastructures section you can find useful
information on current up-and-running Desktop
Grid infrastructures as well as applications that
run on these platforms.
The IDGF portal hosts an interactive forum sec
-
tion where you can discuss Desktop Grid issues
in message boards with the IDGF members and
experts from Desktop Grid projects such as
EDGI and DEGISCO.
For administrative issues, you can contact the
IDGF administration.
Links
International Desktop Grid Federation
http://desktopgridfederation.org
IDGF Membership application
http://desktopgridfederation.org/membership-application
IDGF Experts
http://desktopgridfederation.org/experts
IDGF Downloads
http://desktopgridfederation.org/downloads
IDGF Newsarchive
http://desktopgridfederation.org/news-archive
IDGF Technical documentation
http://desktopgridfederation.org/technical-wiki
IDGF Tutorials
http://desktopgridfederation.org/tutorials
Conferences with IDGF presence and IDGF workshops
http://desktopgridfederation.org/7
Infrastructure and application section
http://desktopgridfederation.org/infrastructures
Interactive IDGF Forum section
http://desktopgridfederation.org/forum
IDGF
administration contact
http://desktopgridfederation.org/contact
3.1.3 Country and local channels

Country Desktop Grid support organisa
-
tions
Alliance Francophone is a community serving
science through Desktop Grid computing using
the BOINC platform. Alliance Francophone
hosts 315.000 volunteers and 520.000 comput
-
ers that participate actively in progressing sci
-
ence. The average computational power of all
participants amounts to 3,5 PetaFLOPS, or 3,5
millions of billions (10 to 15th) of floating opera
-
tions per second.
Rechenkraft.net e.V. supports the development
of the new Distributed Computing-project "Con
-
stellation" in the aerospace domain and asks
interested programmers to participate. Partici
-
pants can support science actively with e.g.
their personal computer, mobile phones or
Playstation 3. They can help in searching for
drugs against cancer and other illnesses. With
-
out prior knowledge, they can take part in re
-
solving highly complex scientific problems. Or
just have a little bit of fun, and have discussions
in the forum/chat and contribute to one of the
many fun projects.
Ibercivis is a citizen computation platform that
allows society to participate on scientific re
-
search in a direct way and in real time. It is an
initiative in Spain that intends to involve in vol
-
unteer computation the maximum number of
citizens. It uses the calculation capacity of the
computer in idle moments to carry out tasks
derived from an investigation project. Ibercivis
brings the citizens closer to leading investiga
-
tions and makes them participant in the scien
-
tific knowledge generation. At the same time, it
provides the scientific community with a power
-
ful calculation tool. The computer turns into an
open window to science, creating a channel for
the direct dialogue between researchers and
society.
Country
Organisation
France
Alliance Francophone
Germany
Rechenkraft e.V.
Spain
Ibercivis
Specialized organisations
The Citizen Cyberscience Centre aims to de
-
velop citizen cyberscience applications targeting
research on humanitarian and development is
-
sues as well as fundamental science, by co-
ordinating collaborative projects between scien
-

28

tists from North and South. It organizes hands-
on workshops for scientists in developing coun
-
tries in order to spread know-how about citizen
cyberscience and its benefits for their research.
The organisation also provides on-line educa
-
tional material about the research challenges
addressed by citizen cyberscience projects, en
-
couraging greater public participation. The cen
-
tre is based on an international partnership.
LHC@home is a volunteer computing pro
-
gramme which enables the user to contribute
idle time on his computer to help physicists de
-
velop and exploit particle accelerators, such as
CERN's Large Hadron Collider. Volunteer com
-
puting makes sense for the LHC. In particular,
volunteer computing is good for tasks which
need a lot of computing power but relatively little
data transfer. In 2004, CERN's IT Department
became interested in evaluating the sort of
technology that is used by volunteer computing
projects like SETI@home. LHC@home became
the overall title for these efforts, and a pro
-
gramme called SixTrack, which simulates parti
-
cles travelling around the LHC to study the sta
-
bility of their orbits, became the first application
to be tested. It was chosen because it can fit on
a single PC and requires relatively little input or
output, but a lot of processing power.
World Community Grid brings people together
from across the globe to create the largest non-
profit computing Grid benefiting humanity. It
does this by pooling surplus computer process
-
ing power. Innovation combined with visionary
scientific research and large-scale volunteerism
can help make the planet smarter. Success de
-
pends on like-minded individuals. World Com
-
munity Grid is making technology available only
to public and not-for-profit organisations to use
in humanitarian research that might otherwise
not be completed due to the high cost of the
computer infrastructure required in the absence
of a public Grid. As part of the commitment to
advancing human welfare, all results are in the
public domain and made public to the global
research community.
Organisation
Location
Citizen Cyber Science
London
LHC@Home
Geneva
World Community Grid
USA
Links
Alliance Francophone
http://www.boinc-af.org/
Citizen Cyber Science
http://citizencyberscience.net/
Ibercivis
http://ibercivis.es
LHC@Home
http://lhcathome.cern.ch/
Rechenkraft e.V.
http://rechenkraft.de/
World Community Grid
http://www.worldcommunitygrid.org/
3.2 Available marketing
and communication
tools
To market Desktop Grids and communicate to
different audiences, you can use a lot of tools.
Remember to choose a tool that fits the audi
-
ence. For instance if you want to communicate
to citizens in the Netherlands on a social media
site, Hyves may be a better choice than Face
-
book.
The tools here are listed with a Volunteer Desk
-
top Grid in mind.
Portal
Message boards
Flyers
Press
Participation in events
Social media
Certification services
3.2.1 Portal
A portal, a website, is the most important com
-
munication tool to keep in touch with volunteers.
You can set up a portal using any of the popular
content management systems. We like to use
Liferay, but Drupal, Joomla, and many others
are just as good (almost). When you are setting

29

up a BOINC project it comes with a web site
out-of-the-box.
Always keep the portal up to date. Write con
-
tent targeted at your audience, preferably in
their own language. Provide information on
what applications you are running and on who
you are (people need to trust you before they
donate computing time).
The portal should also give access to the volun
-
teer's contribution in terms of computing power
to the Grid. Could be this needs to be password
protected. Feedback is important: people want
to know what is done with their computing time.
3.2.2 Message boards
Message boards (Discussion Forums) can be
an important way to communicate with volun
-
teers. All modern content management systems
have discussion forum modules.
A message board in general has to be moder
-
ated. The least you need to do is to answer
questions from the volunteers.
There can also be disadvantages to using mes
-
sage boards: some people tend to dominate
discussions. Savvy volunteers can be very use
-
ful to discuss with on a technical level, but when
laymen see the discussion (in English and not
their native language) they can think “this is not
for me”. Try to clearly separate the discussion
chapters to minimize the risk.
3.2.3 Flyers
Yes, the good old paper flyers. Of course you
put a PDF file in the portal too, but if you meet
someone at an event or meeting, it is always
good to have some flyers at hand for distribu
-
tion.
The flyer should mention the goal of the Desk
-
top Grid, how to participate, and a general url at
a minimum.
3.2.4 Press
Try to build a good relationship with the press.
Do not just send press releases: they will go in
the waste bin if that is all you do - unless your
Desktop Grid does contribute to major scientific
breakthrough.
3.2.5 Participation in events
From a promotion point of view it is very impor
-
tant to regularly organise tutorials and work
-
shops, preferably in joint collaboration with es
-
tablished Grid and Cloud computing confer
-
ences. You can also submit papers in order to
give a presentation about your Desktop Grid
activities at this type of conferences. It will
heighten your visibility and extend your audi
-
ence. If the conference has an exhibition, you
can try to rent a booth. This will give you an
opportunity to talk with interested parties and to
present your Desktop Grid computing project by
means of posters.
3.2.6 Social Media
Many people use social media today. Also new
social media come and “old” social media dis
-
appear. Hence it is useful to choose the right
ones. An important thing to remember is that
you do not really have much influence on what
is happening on social media. Not even with
what happens with your information or your
postings. And it is always good to remember
that even today the majority of people are not
on any social medium.
Video sites, like YouTube or Vimeo can be used
to upload videoʼs about your project, users, or
applications. Vimeo is considered to be more
“professional” while YouTube has a bigger
reach. Do not forget to link back to your own
website.
Twitter is used typically for short news mes
-
sages or to “talk” to other people. There is not a
big volunteer or desktop Grid tweeting commu
-
nity yet.
Profile sites, like FaceBook, LinkedIn, Hyves,
and MySpace can be used to form and partici
-
pate in groups. So they can be useful to come
in contact with other people or new volunteers.
You can also post photos or texts to it. LinkedIN
is more aimed at professionals than the other
sites.
Wikipedia is a kind of encyclopedia, so it is use
-
ful for older projects, or mature applications.
According to Wikipedia policy you should al
-
ways be able to refer to a multitude of scientific
resources, newspapers, etc. And remember
once your text is copied on Wikipedia, it is
Wikipedia that has the copyright. Text for
Wikipedia has to be “new”. And it was good
practice not to write too much about yourself, or
your own company.
As with your own portal, websites or flyers, So
-
cial media have to be maintained. This takes
time, and when your presence is getting more
prominent, it takes more and more time.

30

Links
IDGF on twitter
http://twitter.com//gridfederation
IDGF Vimeo video channel
http://vimeo.com/idgf
Renderfam.fi introduction on YouTube
http://www.youtube.com/watch?v=4VUWrZRCtI8
Rechnenkraft Wikipedia page
http://de.wikipedia.org/wiki/Rechenkraft.net
3.2.7 Certified services
Security is always an issue in volunteer Desktop
Grids. Will the application not harm my ma
-
chine? “Is it not misused?” are often heard
questions. Providing certified services is a way
to handle these issues. IDGF offers certification
services, which allows you to get a certificate for
instance for applications you do provide.
3.3 Experiences in mar
-
keting desktop grids
3.3.1 CityGrid example: AlmereGrid
Background
AlmereGrid was the first CityGrid in the world.
After half a decade, AlmereGrid is internation
-
ally recognized for its pioneering efforts, and
has been and still is, a major partner in several
European Research & Development projects.
AlmereGridʼs main mission is to enable citizens
to support science. The AlmereGrid Volunteer
Computing Grid is running scientific applica
-
tions. People in the city of Almere can donate
their idle computing time.
It also supports scientific experiments in Grid
and Cloud computing that need a distributed
computing test .
AlmereGrid has two main Volunteer Desktop
Grids that each come in two flavours. The most
important Grid is the BOINC based production
Grid. BOINC is the technology used. BOINC is a
very solid and widely used Desktop Grid mid
-
dleware software. Apart from the production
Grid, AlmereGrid also operates a smaller Test
BOINC based Desktop Grid.
The other middleware flavour is XtremWeb.
AlmereGrid also operates a production & test
Grid based on XtremWeb. In total about 3500
computers from volunteers are connected to
AlmereGrid.
At the international level AlmereGrid has as
-
sisted in setting up an International Desktop
Grid Federation (IDGF) for international collabo
-
ration between peers.
Communication message
AlmereGrid was set up to advance science by
the use of innovative technology. The core was,
and still is, setting up a scientific Grid based on
volunteer computing. When over the years, new
technologies come and go, AlmereGrid will in
-
corporate the most useful ones, and contribute
to the research in advanced technologies by
providing test beds.
Communication tools used
As an example we provide the communication
tools we used during 2009-2010.
Event
AlmereGrid organized the Aexact Initiative,
which was titled "From Abacus to Grid". This
initiative was aimed at citizens to involve them
in the use of Desktop Grid computing to solve
scientific questions by organizing an exhibition
with practical examples of Desktop Grid com
-
puting to solve scientific problems and by or
-
ganizing a contest in which participants were
asked to provide a scientific problem which
could be solved by using Desktop Grid comput
-
ing.
The exhibition in the museum “Kunstlokaal Sta
-
tion Muziekwijk” was a first step towards the
vision of bringing more science to the city. The
Station Muziekwijk is an old railway station old
for Almere, the station dates from the mid - that
was partially put out of service.
AlmereGrid also visited children in schools.
Children had to provide difficult questions which
could be solved through Desktop Grid comput
-
ing.
Press release and press contacts
Regional, national and international press brief
-
ings for written and audiovisual press media
have been issued.
Mentions at regional level are important to in
-
form the local citizens about AlmereGrid and the
opportunity to donate computing time.

31

Mentions at national level are of interest for the
same reasons and to show on a national level
what AlmereGrid is doing to help science. This
also helps in attracting new scientific applica
-
tions.
Mentions in international press are important for
the new and existing international collaborations
and the international projects.
Collaboration with other projects
AlmereGrid participated in several European
Desktop Grid computing projects including BE
-
inGRID, a community back-up service for small
businesses; EDGeS, Enabling Desktop Grids
for eScience; EDGI, European Desktop Grid
Initiative, expanding into the Cloud; DEGISCO,
Desktop Grids for International Scientific Col
-
laboration; and Contrail, Open Computing Infra
-
structure for Elastic Services. The objective is
to create a European-wide Desktop Grid opera
-
tors community and to have an impact on the
European politic level.
The AlmereGrid Desktop Experience Work
-
shops aimed to exchange experience between
Desktop Grid operators, partner organisations,
application developers and scientific users.
The scientific part of these workshops concen
-
trated on Desktop Grid Applications for
eScience and eBusiness. Scientific proceedings
have been issued from these workshops.
Links
AlmereGrid
http://almeregrid.nl
Aexact
http://almere2018.org
Aexact press release
http://almere2018.org:8086/almere2018/vmp/repository/n5
4.pdf
Dutch regional television coverage
http://www.omroepflevoland.nl/nieuws/recent?NewsKey=5
479D73C07E37DA0C12575A5004E84D5
3.3.2 Public Grid: EDGeS@Home
Background
EDGeS@Home was set up to support applica
-
tions that also run on the large European serv
-
ice Grid: EGEE (now EGI).
The first application that was used and pro
-
moted was a Nuclear Fusion application called
ISDEP.
EDGeS@Home runs on both BOINC and
XtremWeb middleware. Volunteers can choose.
Communication message
Help solving the world's energy crisis is appeal
-
ing to many citizens. Hence we decided to focus
our initial communication around it.
Communication tools used
A complete communication mix was used:
Launch event
Press release and press contacts
Targeted portal
Flyer
Collaboration with scientists
The
Launch event
was the focal point of the
communication. All other communication tools
were ready at that time. It also offered an oppor
-
tunity to talk to the press about the news. At the
event we invited scientists to talk about the Fu
-
si on r esear ch, exper t s t o t al k about
EDGeS@Home and the technology, and similar
projects to show there is a choice of scientific
projects one can choose from.
A press release was made available in many
languages (EDGeS@Home is international).
Several media were contacted directly in a fol
-
low up.
Links
EDGeS@Home
http://edgesathome.org
Flyer
http://
www.edges-grid.eu:8080/c/document_library/get_file
?p_l_id=29093&folderId=73880&name=DLFE-1618.pdf
Launch event
http://edges-grid.eu/web/userforum/4thtraining
Press release
http://www.edges-grid.eu/web/edges/56

32

3.3.3 A national Desktop Grid: Ibercivis
Background
Ibercivis was developed in Spain with the coop
-
eration of the Institute of Biocomputation and
Physics of Complex Systems at the University
of Zaragoza, CIEMAT, CETA-CIEMAT, Spanish
National Research Council (CSIC) and RedIris.
The project tasks were distributed in different
scientific-technological centres in Spain with the
aim of creating a new platform for science-
based volunteer distributed computing. The pro
-
ject is an European-made development of the
USA-based SETI@home and Berkeley Open
Infrastructure for Network Computing (BOINC)
distributed computing concepts. Ibercivis had
the University of Zaragoza-based distributed
computing project called Zivis as its predeces
-
sor in 2007, and started to work in 2008 in
Spain. Zivis was a pioneer local distributed
computing project funded by the ayuntamiento
(city council) of the city of Zaragoza.
The Ibercivis infrastructure can be used for cal
-
culating applications ranging from fields like nu
-
clear fusion and protein folding to materials
simulations. In July 2009, the Ibercivis platform
was extended to Portugal assuming a transna
-
tional Iberian dimension, following the agree
-
ment to this effect signed by the governments of
both countries in the Luso-Spanish Summit
January 2009 in Zamora, Spain. Portuguese
institutions affiliated with Ibercivis include the
Ministry of Science, Technology and Higher
Education, the Centre for Neuroscience and
Cell Biology at the University of Coimbra, and
the LIP experimental high energy physics labo
-
ratory.
Ibercivis is not a temporary project - unlike
many BOINC-based volunteer computing pro
-
jects, so it will be possible to submit applications
indefinitely. In addition, Ibercivis is designed to
run not just one but several applications belong
-
ing to different disciplines, in a manner similar to
the IBM-funded World Community Grid.
Communication message
Ibercivis is a joint scientific collaboration of the
Portuguese and Spanish governments, but is
open to citizens and scientists of other countries
outside the Iberian Peninsula who wish to join
the initiative. It uses the calculation capacity of
the computer in idle moments to carry out tasks
derived from an investigation project.
Ibercivis brings the citizens closer to leading
investigations and makes them participant in the
generation of scientific knowledge. At the same
time, it provides the scientific community with a
powerful calculation tool. The computer turns
into an open window to science, creating a
channel for the direct dialogue between re
-
searchers and society.
Communication tools used
Ibercivis has launched a monthly publication on
research carried out in Ibercivis. The format of
these publications is an interview with a princi
-
pal researcher for each application.
The Ibercivis project has Teaching Units whose
main purpose consists in bringing scientific re
-
search to secondary schools.
The project organises workshops at regular
times. For example, researchers and techni
-
cians from different schools as the universities
of Extremadura, Granada, Autónoma de Madrid
and the CSIC, among others, attended on 8th
and 9th September 2010 the I Ibercivis Work
-
shop, held at the Institute of Biocomputing and
Physics System Complex (BIFI). The purpose of
this meeting was to bring positions between the
three pillars on which it is based Ibercivis: re
-
search, development and communication, creat
-
ing workshops around different development
proposals and communication.
Links
Ibercivis
http://www.ibercivis.es
Didactical material
http://www.ibercivis.es/index.php?module=public&section=
channels&action=view&id_channel=4&id_subchannel=38
Monthly publication
http://www.ibercivis.es/content/modules/webboard/channels/
files/00000064_df6809f06827594bbc84e7c0dda36c4f.pdf
Ibercivis Workshop
http://www.ibercivis.es/index.php?module=public&section=
channels&action=view&id_channel=4&id_subchannel=135

33

4. Legal and organisa
-
tional information
The following matrix summarizes the most im
-
portant topics among the key players of Desk
-
top Grids in details.
This matrix allows us to discover the possible
issues in complex cases when the policies must
be harmonized e.g. with Service Grids. How
-
ever, please note that in academic or university
environments several roles are merged into one
key player quite frequently. For example, the
end-user provides the data and the application
with his/her own data. It simplifies the legal
situation and organisation issues; the matrix will
have only 3 rows and columns; resource own
-
ers, Desktop Grid service provider and the
merged data owners/application provided/data
provider/end-user. This is the case, when the IT
department creates a
public volunteer Desktop
Grid
for the research department in order to run
their own application with the own data of the
research department.
In the most simplest case when the target is to
establish a
local Desktop Grid
, which is oper
-
ated locally and used inside a research lab by
itself using their own data and application pro
-
vided by the local developers from the local re
-
search lab; the legal issues can be maximally
eliminated.
The matrix intends to cover the situation when
the desktop grid or any other service is provided
for free.
resource
owners
data own
-
ers
Desktop
Grid service
providers
application
provider
data pro
-
vider (stor
-
age)
end-user
resource
owners
(through
granting ac
-
cess to DG
service pro
-
vider)
grant access
to his/her
own re
-
sources by
joining a DG
project
(through
granting ac
-
cess to DG
service pro
-
vider)
(through
granting ac
-
cess to DG
service pro
-
vider)
(through
granting ac
-
cess to DG
service pro
-
vider)
data owners
(through vali
-
dation proc
-
ess of EADM)
grant access
to (ano
-
nymized) data
in production
grant access
to (ano
-
nymized) data
during devel
-
opment
store (ano
-
nymized) data
in storage of
DG provider
allow the
exploitation
of data
Desktop Grid
service pro
-
viders
No warranty
but validate
applications
with EADM,
and protect
resource
owners pri
-
vacy
No warranty
but apply
best
practices/
technologies
and recom
-
mend ano
-
nymization
provide sup
-
port and host
the application
grant access
during devel
-
opment and in
production
provide ac
-
cess with
SLA or with
best effort
base with
privacy
statement
application
providers
(through vali
-
dation proc
-
ess of EADM)
accept the
data owner
policy
collaborate in
testing and
validation
grant access
to stored data
use Grid-
friendly SW
license (or
proprietary
software)
and hold
IPR
data provider
(storage)
(through vali
-
dation proc
-
ess of EADM)
ensure pri
-
vacy and con
-
fidentiality
ensure secure
access from
Grid
ensure secure
access from
apps
grant access
to stored
data
end-users
(through ac
-
cepted AUP)
accept the
data owner
policy
accept AUP
from the grid
service pro
-
vider
accept SW
licensing
policies (or
proprietary
software)
accept policy
of data pro
-
vider

34

Links
Grid and Cloud Computing: A Business Perspective on
Technology and Applications
http://www.springer.com/business+%26+management/busi
ness+information+systems/book/978-3-642-05192-0
(chapter 7)
Memorandum of Understanding between EGEE and
EDGeS
http://www.edges-grid.eu/c/document_library/get_file?folde
rId=36519&name=DLFE-1212.pdf
Memorandum of Understanding on Collaboration (be
-
tween EDGeS and SEE-GRID-SCI)
http://edges-grid.eu/web/edges/37
Concerning the legal issues we focus on the
key players with special liabilities and rights;
end-user of Desktop Grids, Desktop Grid pro
-
viders, data and resource owners of Desktop
Grids.
4.1 Privacy issues
In order to ensure data anonymization data pro
-
tection best practices and standards must be
followed by several key players. It includes the
European Directive 95/46/EC on the protection
of individuals, e.g. resource providers for volun
-
teer Desktop Grids, with regard to the process
-
ing of their personal data, and concerning the
data privacy the various service providers and
data owners must follow the national or local
guidelines as well.
Three major cases can be distinguished:
Between End-user / Desktop Grid service
providers: Desktop Grid service providers are
responsible to handle the private data (email
addresses, names, IP addresses) of end users
according to the above described policies.
Between Resource owner / Desktop Grid
service provider: Desktop Grid service providers
are responsible to handle the private data (e.g.
email addresses, names, IP addresses) of re
-
source owners in case of volunteer Desktop
Grids according to the general policies.
Between Data owners / data provider (stor
-
age) or any further players: As a thumb rule,
data owners must be responsible for anonymi
-
zation of the input data before delivering them
to the data provider or any other players later.
4.2 Licensing
The Desktop Grid middleware solutions and
tools, which are supported and recommended
by IDGF, are open-source and free software
packages mainly from the EDGI and DEGISCO
project partners.
The ported and available applications from the
repositories. and data from the data providers
must have Grid-friendly licensing as well. How
-
ever, the new applications can be either proprie
-
tary or open-source ones. For local usage the
proprietary application code is not a major bar
-
rier but volunteer Grids mostly require open-
source applications in order to build and keep
the trust.
4.3 Misuse
The candidate applications which will use a vol
-
unteer Desktop Grid system for their execution,
must be validated using a test infrastructure
based on EADM. In case of local desktop Grids,
this phase is optional. The validation process
ensures that the infrastructure could not be
used for illegal computations. The introduction
of Acceptable Use Policy helps ensure this re
-
quirement, too.
4.4 Warranties
The validation process in the EADM does not
allow harmful jobs, which could destroy part or
the whole infrastructure. In a part of the valida
-
tion process the application is tested in a limited
test environment, not the whole infrastructure.
Despite of the best efforts in the EADM, due to
the nature of the applied software stack (e.g.
using BOINC distribution) which are provided
without warranty and 'as is', the appropriate dis
-
claimer must be placed on each document,
software component, and service entry points.
An example in case of software:
Disclaimer of Warranty
THIS SOFTWARE AND THE ACCOMPANYING
FILES ARE DISTRIBUTED "AS IS" AND
WITHOUT WARRANTIES OF ANY KIND, AND
THE IMPLIED WARRANTIES OF MERCHANT
-
ABILITY OR FITNESS FOR A PARTICULAR
PURPOSE ARE HEREBY DISCLAIMED. ANY
STATUTORY WARRANTY OF NON-
INFRINGEMENT IS ALSO DISCLAIMED.

35

4.5 Security based on
Grid certificates
Obtaining, using, and renewing a certificate by
an average Grid user is very often a cumber
-
some process especially for scientific end-users
without solid IT skills, and in some ICPC coun
-
tries (where the CA does not functioning) as
well. The solution of most Desktop Grids follows
an approach which lowers the barrier to entry in
using Desktop Grids and also allows the user to
not posses a certificate. Only in some cases the
user must have valid certificate issued by a Cer
-
tificate Authority:
obviously, when submitting jobs from a Serv
-
ice Grid through the 3G Bridge by a Service
Grid end-user
if the end-user and the application provider
are the same, and he/she intends to execute
the jobs on a Service Grid through the 3G
bridge
4.6 IPR
As a recommendation; the intellectual property
rights generated by a Party under a DG collabo
-
ration shall be the property of that Party who
shall be free to protect, transfer and use such
Intellectual Property Rights as it deems fit. Not
-
withstanding the foregoing each Party shall
grant the other a non exclusive royalty free,
perpetual license to use the Intellectual Property
Rights generated by it under the DG collabora
-
tion for use within its project or for the exploita
-
tion the results thereof. Such license shall in
-
clude the right to sublicense the entities in
-
volved in the project.
4.7 Interoperation with
Service Grids
Since the number of service grids in production
is rather limited and the service grids have vari
-
ous policies; connecting the new Service Grid to
a Desktop Grid requires negotiation between
the Grid operators. Two examples can be taken
into consideration: the Memorandum of Under
-
standings signed between the EDGeS and the
SEE-GRID-SCI and the EGEE-III projects.
The major topics that must be covered are the
following with some recommended resolutions:
responsibility: Each operator will maintain full
responsibility for its own activities
IPR: proprietary rights relating to the results
of activities undertaken by the collaborators un
-
der the MoU shall be jointly held by the projects
warranty: no warranty
licenses and access rights: a project grants
the other project a free, irrevocable and perpet
-
ual license to use such rights for joint IPR

36

5. Green Desktop Grids
Desktop Grids are the real Green Grids
Key advantage of Desktop Grids is the minimal
power density compared to conventional data
centres. Typically, PCs participating in Desktop
Grids in Europe are not hosted in air-
conditioned environments. Without the energy
burden of air-conditions, Desktop Grid are in
-
trinsically “greener” than data centre based clus
-
ters and thereof built Service-Grids.
The term “Green-Desktop Grids” has been in
-
troduced along with the first hype on Green-
Computing”. Here we want to discuss why
wise
use of energy

is highly relevant for Desktop
Grids and how energy saving methodologies
can be applied.
5.1 Relevance of Green-
Desktop Grids
5.1.1 Responsibility for real costs: Wise or
Waste use of energy
Desktop Grids gather donor machines up into
the 100-thousands per project.
Let us assume the average modern PC uses
about 300 Watts full power. That individual
power consumption times those 100.000 ma
-
chines result in tremendous 30MW of power
consumption. A typical project may consist of
200.000 jobs with an average runtime of 2
hours. Also for example chose a typical end
customer price of 25 Eurocts/kWh for electricity:
200.000 Jobs x 2 hours x 300 Watts = 120MWh
for one campaign
120MWh x 25cts/kWh = 30.000 Euro contrib
-
uted by donors for one campaign or 15cts/job.
Now, asking contributors for their machines and
compute time actually means asking them to
pay for energy – Desktop Grid computing is not
free of costs.
Operators and users (=scientists) of Deskto
Grids have the responsibility to make wise use
of this energy.
5.1.2 Selection of Green IT arguments: CO2
footprint and the energy mix
One of the originally formulated Green IT aims
was to reduce CO2 footprint of IT activities. A
broader formulation targets the reduction of en
-
ergy consumption in general and especially
thermal emissions in metropolitan areas, both
impacting
a) global climate
b) local (micro) climate and
c) human quality of life.
Production of CO2 and accordingly the reduc
-
tion of CO2 footprint are difficult to measure
from the perspective of a concrete IT activity like
computation. Even if energy consumption as
such is accounted for, it depends on the local
energy mix how much CO2 this is equivalent to.
The electrical energy mix, the combination of
electrical energy sources, depends on national
specifics, see table. E.g. Denmark produces
25% of its electricity consumption from wind –
sometimes up to 150% (when strong winds pro
-
duce more electricity than Denmark needs)
causing negative energy prices at the spot mar
-
ket. Neither the single PC owner who contrib
-
utes compute time to a scientific project nor the
operator of the Desktop Grid server that offers
the respective workload item is in control of the
energy source used to produce electricity con
-
sumed by the computational effort.
5.1.3 € - a metric for Green IT success
In order to measure the effectiveness of energy
saving policies and methods, we need to intro
-
duce a metric that can be “metered”. The obvi
-
ous advantage of “kWh” as the base metric for
Green IT is the simplicity of measurement: elec
-
tricity is metered everywhere. Different from
computer centres and conventional Service-
Grids, policies and methods are applied and
executed in Desktop Grids only by the volunteer
effort of the resource contributor.
Example: a donor located in France will make
use of energy produced by renewals + nuclear
= ~90%. Only 10% of his or her energy con
-
sumption is connected to CO2 production. CO2
footprint for computation in France is mostly
irrelevant while we agree that saving energy is
nevertheless of high importance.
As success metric for Green IT, the translation
into cost, into money, is helpful to connect to
business considerations and propel motivation.

37

With € (for kWh) as metric, contributors can
relate their choice of workload and policy-
compliance to the personal electricity bill: Green
Desktop Grids help the planet and your budget!
5.1.4 The eleventh commandment: You shall
not waste energy
The Green-Methodologies describe a number of
measures to make best possible use of energy.
The energy used for computation does not al
-
ways result in waste-heat but sometimes in well
received dual use when heating is based on
electricity anyway and actual ambient tempera
-
ture makes additional heating necessary.
In most situations, consumed energy equals
waste-heat and good consideration should be
taken to assure responsible use of it.
5.2 Green Desktop Grid
Methodologies
The short-list of 7 methodologies is a collection
of best practices, techniques and policies.
Some of these methodologies are of technical
nature, some of organisational. Some of them
are readily implemented, some still object of
research.
Organisational methodologies:
Exploitation of natural ambient condi
-
tions
Cool strategy: avoid air-condition use
Technological methodologies
Energy profiling of applications
Ambient metrics based Green opti
-
mization
Time-of-day and weather dependent
energy tariffs
CPU speed steps
Switch off unused machines
All those recommended methodologies are
based on the fundamental principles of Desktop
Grids: the volunteer, the donor stays in control
of her/his resources. It is up to the donor to ac
-
cept advice on e.g. energy-aware configura
-
tions.
We do not even discuss potential remote-control
of donor machines and do not intend to provide
technology means to centrally enforce certain
settings or behaviour.
The implementation of the Green Desktop Grid
methodologies takes place by information pre
-
sented to the donor with the invitation to chose
her/his preferred options to optimize energy-
efficiency.
The only exception are Desktop Grids that are
centrally managed, like e.g. “The University of
Westminster Local Desktop Grid” described be
-
low. Here the owner, the university, is the “vol
-
unteer” and implements centralized manage
-
ment means for all their resources, also capable
to adjust Desktop Grid-Client configurations.
Companies or institutions considering to con
-
tribute groups of resources will use their
Windows-clients management facilities to
achieve this level of control.
In consequence, we do not see the need (yet -
unless you educate us differently) to provide
remote-control of donor machines as native fea
-
ture of Desktop Grid technology.
5.2.1 The balance between complexity and
Green features
Managing a Desktop Grid in a “greener” way
adds complexity to the already not trivial base
functionality. Adding complexity is acceptable
only below a certain threshold: The aim to save
energy shall not cause technological instabilities
or excess operational difficulties. To which ex
-
tend these difficulties occur, depends primarily
on the used Desktop Grid technology. Middle
-
ware like BOINC does not allow to steer and
control the donor's desktop machine. Conse
-
quently, switching off PCs must make use of
extensions or external techniques which results
in additional complexity. Other Desktop Grid
middleware like OurGrid has this feature built-in
and can even select between suspend ad hi
-
bernate modes.
When planning a new Desktop Grid, the appro
-
priate balance between preferred features and
resulting complexity will have to be considered.
5.2.2 Powerful organisational methodologies
Organisational methodologies can be imple
-
mented mostly independent of the chosen
technology, still they provide excellent results.
Both organisational methodologies make use of
the fact, that the Desktop Grid donor frequently

38

is sufficiently educated and interested to follow
instructions.
Exploitation of natural ambient conditions
When advertising a new scientific challenge,
a new project, one may add recommendations
to donors from climate-wise suitable regions.
Different locations yield different climates
Kazakhstan, Amaty
http://worldweather.wmo.int/070/c00152.htm
Russia, Moscow
http://worldweather.wmo.int/107/c00206.htm
Hungary, Budapest
http://worldweather.wmo.int/017/c00060.htm
Denmark, Copenhagen
http://worldweather.wmo.int/173/c00190.htm
Spain, Zaragoza
http://worldweather.wmo.int/083/c01240.htm
By this targeted recruiting, still the donor de
-
cides what to subscribe to. By adding informa
-
tion on assumed power consumption per com
-
pute task and resulting heat production, donors
can take informed decision on whether to ac
-
cept this workload type – and many of them will.
Cool strategy: avoid air-condition use
Desktop Grids are the real Green Grids: lower
energy density than clusters results in less en
-
ergy wasted for cooling. However, this may not
longer be true if air-conditions are used to as
-
sure proper operation of Desktops. Principles of
cooling: Energy consumption by air-conditions
range from 30% to >200% of the energy dissi
-
pated by the IT device (payload), depending on
the cool-reservoir temperature the heat pump
can utilize to get rid of the heat. The prime ad
-
vice to configure Desktop Grids: avoid air-
condition use! Selection criteria for the “maxi
-
mum temperature” as described above could be
that temperature which would just not yet trigger
the start of the local air-condition. Applications
could be chosen by the contributor which are
indicated to produce a certain heat/time only
(compare “Energy profiling of applications”), by
keeping the ambient temperature in acceptable
range.
Energy profiling of applications
To understand the potential of energy profiling,
we need to consider that energy consumption
and by this, heat production of computers is
massively influenced by the workload.
As listed in great detail for thousands of com
-
puter products by Energy-Star, typically the en
-
ergy consumption varies for a factor of about 2
to 4 between idle and fully loaded. In other
words: when a PC is idling with some 50 Watts,
Desktop Grid activity may induce beyond 200
Watts of energy consumption and heat genera
-
tion, making use of all cores in the box.
A pure CPU bound application would possibly
be able to run this CPU (and possibly even all 4
or 6 or 12 cores) to 100%, pushing the total
power consumption to the maximum.
Another application will work with some I/O to
RAM or HDD, effectively keeping the power and
temperature of the PC lower.
Different applications and codes consume more
or less CPU at any given time, resulting in dif
-
ferent energy consumption per time interval;
they behave differently in raising machine and
ambient temperature. Accordingly applications
can be classified with a heat index as +, ++ and
+++ for example. The +++ index marks an ap
-
plication that makes maximum use of a given
machine, is raising its temperature, but finishes
the computation quickly. This behaviour may
total in less “energy consumed/computation”
than the application which creates less heat/
time. Still heat/time is an important parameter
from a green operations point of view.
Time-of-day and weather dependent energy
tariffs
The value of electrical energy is usually chang
-
ing according to the conditions of generation as
well as by changing consumption. Accordingly,
the tariffs for electricity are changing.
A typical profile in Central Europe would be
16cts/kWh during night, 23cts/kWh in daytime
and 80cts at noon +/- 1hour.(private customer
prices). Other regions with different climate,
style-of living (=time to cook) and different en
-
ergy production mix have completely different
time-of-day profiles.
Typically, the peak usage is covered by Oil and
Gas turbine power plants. Reducing computa
-
tional workload during those peak times save
CO2 directly.

39

In the example given above, Denmark, the wind
produced electricity is predictable for a couple
of hours (up to ~1day) ahead. An accordingly
updated Desktop Grid project website could
advice the donors to “contribute now” as his
energy price would be low at this point in time.
To improve the energy cost situation and to take
advantage of excess Green electricity, advice
could be given to contributors how to configure
their Desktop Grid-Clients to prefer workload
during low tariff times.
5.3 Implementation
Here we will list a growing number of implemen
-
tation examples of Green Desktop Grid strate
-
gies.
5.3.1 The University of Westminster Local
Desktop Grid
The University of Westminster Local Desktop
Grid (WLDG) connects laboratory PCs of the
University of Westminster (London, UK) into a
BOINC based Desktop Grid infrastructure. Over
1600 of these machines are connected to the
WLDG. The installed desktop Grid software is
the Local SZTAKI Desktop Grid package. This is
a modified version of the original BOINC instal
-
lation that focuses on the requirements of non-
public DG infrastructures. In these scenarios the
resources are controlled centrally (in the case of
the WLDG by the central computing services of
the university), and there is no need for a public
Website or the credit system to attract donors.
BOINC client configuration: the DG system
works in parallel with the energy saving soft
-
ware introduced by the university. This solution
automatically shuts the PCs down if there was
no user activity on them for a set period of time
(typically 20 minutes). Therefore, if students log
out and there are no DG jobs running on the
computer within the next 20 minutes, then the
computer automatically shuts down and can
only be restarted manually by the next user.
However, if there are DG work units executed
on the computer then it remains active until the
continuous load of these WUs is guaranteed (or
someone shuts the PC down manually).
5.3.2 OurGrid, the Brazil Desktop Grid saves
energy
In their presentation “On the Impact of Energy-
saving Strategies in Opportunistic Grids” Le
-
sandro Ponciano and Francisco Brasileiro de
-
scribe their research on finding the perfect bal
-
ance between “Decrease Power Consumption”
and “Increase Latency to wake-up” by leaving
clients idle or switching them to Standby or Hi
-
bernate.
The OurGrid middleware supports levels of cli
-
ent control that are unique for Desktop Grids.
See also the fully detailed paper: “On the Im
-
pact of Energy-saving Strategies in Opportunis
-
tic Grids” .
5.3.3 BOINC: Heat and energy considera
-
tions
The BOINC Wiki:
http://boinc.berkeley.edu/wiki/Heat_and_energy
_considerations
provides some considerations on energy related
adjustments available for the contributor.
Three concrete advices are:
1.
Set your preferences to allow BOINC to
compute while your computer is in use, and
turn your computer off when it's not in use.
2.
Use your computer's power-management
features to turn off your monitor when it's not
in use, or to enter a low-power mode.
3.
If your electricity costs vary according to time
of day, set your preferences so that BOINC
computes only during periods of low electricity
costs.
Advice #3 is congruent with our “time-of-day
tariff” methodology, the according BOINC client
configuration is shown in the next chapter.
Advice #2 is a clear “no-brainer”: Yes, you
should configure your screen-saver as black/
blanc screen and power options to switch off the
screen after some minutes when no interactive
usage takes place.
Advice #1 may need some re-consideration:
understanding that energy consumption is com
-
posed of a base (~50%) and a usage related
(~50%) component, the advice is correct: the
total energy consumption is minimized when
Desktop Grid-Clients are allowed to use the
machine only when interactive usage take place
concurrently. It should be noted that this advice
spoils the Desktop Grid “logistics base assump
-
tion”: the point of Desktop Grids is to collect
CPU cycles when the local user does not need
them – which implies that the Desktop Grid-
Client is enabled to use the machine exactly

40

when the interactive user is gone. When concur
-
rent usage is enforced, the Desktop Grid work
-
load may not be able to allocate sufficient re
-
sources – it will take longer to complete and
may even be interrupted more often, potentially
further reducing efficiency by going through
more checkpoint-restart cycles.
We recommend to consider the actual project
type and targeted donor group before propagat
-
ing advice #1.
Energy related BOINC client configuration
The BOINC Preferences can be used to control
the behaviour of the BOINC client with regards
to energy-consumption.
The Boinc Wiki
(http://boinc.berkeley.edu/wiki/Preferences) ex
-
plains them quite well:
You can specify preferences that limit when and
how BOINC uses your computers. There are
two kinds of preferences:
General preferences
These apply to all projects. Example: whether
BOINC should compute while you are at your
computer.
Project preferences
These apply only to one project. Example: the
colour scheme used in the SETI@home
screensaver.
Adjustment of preferences can be made via
Project web site
BOINC Manager, Advanced/Preferences menu
e d i t i n g t h e l o c a l p r e f e r e n c e s fi l e
“global_prefs_override.xml”
global_prefs_override.xml
From
http://boinc.berkeley.edu/wiki/Global_prefs_over
ride.xml :
Local preferences are stored in a 'preferences
override file'. This file is read by the core client
after it reads the preferences from the server,
and it overrides those preferences. You can
modify the preferences override file directly (in a
text editor) if you like (for example, if you do not
have the BOINC Manager installed).
After editing the preferences file, go to the Ad
-
vanced View in BOINC manager and click Ad
-
vanced -> Read local prefs file to make your
new preferences take effect. If you do not have
BOINC manager installed, restart the client or
do boinccmd --read_global_prefs_override from
a command line to have your new preferences
take effect.
The preferences override file is named
`global_prefs_override.xml` and it is located in
the BOINC data directory. Its structure is as fol
-
lows: (… indicates omissions)
<global_preferences>
...
<start_hour>%f</start_hour>
<end_hour>%f</end_hour>
...
<max_cpus>%d</max_cpus>
...
<cpu_usage_limit>%f</cpu_usage_limit>
<day_prefs>
<day_of_week>%d</day_of_week>
<start_hour>%.02f</start_hour>
<end_hour>%.02f</end_hour>

<net_start_hour>%.02f</net_start_hour>
<net_end_hour>%.02f</net_end_hour>
</day_prefs>
</global_preferences>
Of interest for energy-aware configuration are
the following parameters:
A) Maximum number of processors to use
<max_cpus>%d</max_cpus>
On a multiprocessor, this limits the number of
processors that BOINC will use. As of 6.1, the
option works as a % selector of total available
processor cores.
B) Use at most X% of CPU time
<cpu_usage_limit>%f</cpu_usage_limit>
If you specify 50%, BOINC will compute only
every other second. This reduces the heat out
-
put and energy usage of your CPU chip.
Parameters A) and B) can be used to control
the heat dissipated from the Desktop Grid-
Client, useful to implement methodologies on
Cool strategy: avoid air-condition use
fine granular control on how much heat is dissi
-
pated – allows to keep off airconditions.
Energy profiling of applications

41

==> execute a +++ application in a raised ambi
-
ent temperature environment by reducing the
amount of CPU available.
Ambient metrics based Green optimization
execute workload although ambient is high.
C) Working hours
The parameters:
<start_hour>%f</start_hour>
<end_hour>%f</end_hour>
allow to define when the Desktop Grid-Client is
executing workload.
Parameter C) can be used to reflect time-of-day
dependent energy tariffs: allow Grid workload
only during low cost times.
The parameters section <day_prefs> provides
ever finer granularity of control, inclusive defini
-
tion of working days.
<day_prefs>
<day_of_week>%d</day_of_week>
<start_hour>%.02f</start_hour>
<end_hour>%.02f</end_hour>

<net_start_hour>%.02f</net_start_hour>
<net_end_hour>%.02f</net_end_hour>
</day_prefs>
Caveat: the discussion tab of the referenced
BOINC Wiki points to inconsistencies in using
<day_prefs>. Before recommending the use of
<day_prefs> to the donors, a verification test
should take place with the SW versions used for
your concrete project.
5.3.3 BOINC client configuration for time-of-
Day energy tariffs
BOINC volunteers can exactly tell BOINC when
and how their computer can be used.
Links
Energy-Star
http://www.energystar.gov/index.cfm?fuseaction=find_a_pr
oduct.showProductGroup&pgw_code=CO
Lesandro Ponciano, Francisco Brasileiro, On the Impact of
Energy-saving Strategies in Opportunistic Grids,
http://2446993791577007343-a-1802744773732722657-s-
sites.googlegroups.com/site/lesandrops/papers/E2GC2-20
10-camera-ready-ID8.pdf?attachauth=ANoY7cprWV-3oGB
guOuL2TyvHp5cz0GxTFDkYZeKjUz0aLepmsj3Zbh8Wmn
ql2sRajnhemRrI129StkU0356hKE2LXyADZaJzEqvod2ON
XHYAaouQgvHoppoo5MVQ7JeEnMxbA8Z8mT91Rc1V8U
TJZ_aa6bDmkil8rwgJEP0Tbh_l87DozAnwQwjbUrWtYr57
QCNQWurUSjpmq_4fROeUHjMqc-3fq9rSQbkXo09wUDD
WjKx5y9e4ws%3D&attredirects=0
Lesandro Ponciano, Francisco Brasileiro, On the Impact of
Energy-saving Strategies in Opportunistic Grids,
http://www.ens-lyon.fr/LIP/RESO/e2gc2_2010/slides/e2gc2
_lponciano.pdf
Comission, European. GERMANY – Energy Mix Fact
Sheet. [Online] 01 01 2007
http://ec.europa.eu/energy/energy_policy/doc/factsheets/m
ix/mix_de_en.pdf
Commission, European. FRANCE – Energy Mix Fact
Sheet. [Online] 01 01 2007
http://ec.europa.eu/energy/energy_policy/doc/factsheets/m
ix/mix_fr_en.pdf
Comission, European. DENMARK – Energy Mix Fact
Sheet. [Online]

42

http://ec.europa.eu/energy/energy_policy/doc/factsheets/m
ix/mix_dk_en.pdf
Spot, Nord Pool. Nord Pool Spot implements negative
price floor in Elspot from October 2009. Press release.
[Online] 04 02 2009.
http://www.nordpoolspot.com/Market_Information/Exchang
e-information/No162009-Nord-Pool-Spot-implements-nega
tive-price-floor-in-Elspot-from-October-2009-/

43

6 Glossary
AUP
Acceptable Use Policy
CO
Confidential
CONDMAT
International Conference
Modern Problems of Con
-
densed Matter
DEGISCO
Desktop Grids for Interna
-
tional Scientific Collaboration
DG
Desktop Grid
EADM
EDGeS Application Develop
-
ment Methodology
EC
European Commission
EDGeS
Enabling Desktop Grids for e-
Science
EGI
European Grid Initiative
EGITF
European Grid Initiative Tech
-
nical Forum
e-IRG
e-Infrastructure Reflection
Group
GNU
GNU
NGI
National Grid Initiative
ICPC
International Cooperation
Partner Countries
IDGF
International Desktop Grid
Federation
IPR
Intellectual Property Right
ISGC
International Symposium on
Grids and Clouds
OGF
Open Grid Forum
PU
Public
RSS
Remote Site Syndication
SG
Service Grid
SLA
Service Level Agreement
WP
Workpackage

44

A special thanks to all the volunteers
around the world that use Desktop
Grids to donate idle computing time and
thus giving science a helping hand.