Energy Efficient Data Centers: Third International Workshop, E2DC 2014, Cambridge, UK, June 10, 2014, Revised Selected Papers

173 Pages·2015·22.145 MB·English

by  Sonja Klingert, Marta Chinnici, Milagros Rey Porto (eds.)

Sonja Klingert
Marta Chinnici
Milagros Rey Porto (Eds.)
5 Energy Efficient
4
9
8

S
Data Centers
C
N
L
2
Third International Workshop, E DC 2014
Cambridge, UK, June 10, 2014
Revised Selected Papers
123
Lecture Notes in Computer Science 8945
Commenced Publication in 1973
Founding and Former Series Editors:
Gerhard Goos, Juris Hartmanis, and Jan van Leeuwen
Editorial Board
David Hutchison
Lancaster University, Lancaster, UK
Takeo Kanade
Carnegie Mellon University, Pittsburgh, PA, USA
Josef Kittler
University of Surrey, Guildford, UK
Jon M. Kleinberg
Cornell University, Ithaca, NY, USA
Friedemann Mattern
ETH Zurich, Zürich, Switzerland
John C. Mitchell
Stanford University, Stanford, CA, USA
Moni Naor
Weizmann Institute of Science, Rehovot, Israel
C. Pandu Rangan
Indian Institute of Technology, Madras, India
Bernhard Steffen
TU Dortmund University, Dortmund, Germany
Demetri Terzopoulos
University of California, Los Angeles, CA, USA
Doug Tygar
University of California, Berkeley, CA, USA
Gerhard Weikum
Max Planck Institute for Informatics, Saarbrücken, Germany
More information about this series at http://www.springer.com/series/7411
Sonja Klingert Marta Chinnici
(cid:129)
Milagros Rey Porto (Eds.)
Energy Efficient
Data Centers
2
Third International Workshop, E DC 2014
Cambridge, UK, June 10, 2014
Revised Selected Papers
123
Editors
SonjaKlingert MilagrosReyPorto
Universityof Mannheim GasNaturalFenosa,TechnologicalProjects
Mannheim Barcelona
Germany Spain
Marta Chinnici
ICT-Division Italian NationalAgency
forNew TechnologiesEnergy
Rome
Italy
ISSN 0302-9743 ISSN 1611-3349 (electronic)
Lecture Notesin ComputerScience
ISBN 978-3-319-15785-6 ISBN 978-3-319-15786-3 (eBook)
DOI 10.1007/978-3-319-15786-3
LibraryofCongressControlNumber:2015932621
LNCSSublibrary:SL5–ComputerCommunicationNetworksandTelecommunications
SpringerChamHeidelbergNewYorkDordrechtLondon
©SpringerInternationalPublishingSwitzerland2015
Thisworkissubjecttocopyright.AllrightsarereservedbythePublisher,whetherthewholeorpartofthe
material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,
broadcasting, reproduction on microfilms or in any other physical way, and transmission or information
storageandretrieval,electronicadaptation,computersoftware,orbysimilarordissimilarmethodologynow
knownorhereafterdeveloped.
Theuseofgeneraldescriptivenames,registerednames,trademarks,servicemarks,etc.inthispublication
doesnotimply,evenintheabsenceofaspecificstatement,thatsuchnamesareexemptfromtherelevant
protectivelawsandregulationsandthereforefreeforgeneraluse.
Thepublisher,theauthorsandtheeditorsaresafetoassumethattheadviceandinformationinthisbookare
believedtobetrueandaccurateatthedateofpublication.Neitherthepublishernortheauthorsortheeditors
give a warranty, express or implied, with respect to the material contained herein or for any errors or
omissionsthatmayhavebeenmade.
Printedonacid-freepaper
SpringerInternationalPublishingAGSwitzerlandispartofSpringerScience+BusinessMedia
(www.springer.com)
Preface
The spread of new Data Center facilities around the world is an accompanying phe-
nomenon of the twenty-first century always-on-connected-everywhere lifestyle. New
Data Centers are being opened up both in Iceland and in urban conglomerations like
Frankfurt in Germany and New York in the USA. Legacy Data Centers are being
enlarged and continually updated with new equipment and management frameworks.
Globally, this evolution results in an ever-increasing energy consumption of Data
Centers, stipulating climate change and human impact on the earth’s surface. The
conceptofenergyefficiencyinDataCentersthatonlyafewyearsagowasrestrictedto
enhancing IT equipment and cooling, is today addressed to a variety of system level
technologies and associated services that will improve energy and environmental
performance.TheattentionisfurthermorefocusedonsoftwarerunninginDataCenters
and the way that workload is being processed. However, as power consumers Data
Centers additionally need to be viewed as part of a greater system. This applies for
instancetotheroleDataCentersplayinthecontextofSmartCities.DataCentersform
animportantpartofcitiesandplayaleadingroleasanenablerofcityservices,butthey
arealsohugepowerconsumers.ThispertainsalsotoDataCentersasmajorplayersin
thepowergrid.ReducingthecarbonfootprintofDataCentersworldwideisthereforea
huge challenge considering the pressure of rocketing data amounts. However, prom-
ising starting points can be found both in academic and commercial research projects,
as the International Workshop on Energy Efficient Data Centers E2DC 2014 was able
toshowonceagain.Forthethirdtime,researchersfromaroundtheworldmetinorder
to commonly advance knowledge and experience of reducing Data Center energy and
power consumption and aligning Data Center power profiles to the availability of
renewable power resources or constraints from the power grid. The workshop was
collocated with the ACM SIGCOMM e-Energy 2014 conference on June 10, 2014 in
Cambridge, UK and organized by the EU FP7 project DC4Cities1.
These proceedings of the workshop give an account on high quality papers from a
huge range of relevant technologies within Data Centers as well as regarding the
interaction of a Data Center with its environment aimed at saving energy and inte-
grating renewable energy sources.
Thefirstpartoftheproceedingscontainsfourpapersdevotedtoenergyoptimization
algorithms and models. Yi and Singh proposed a greedy algorithm capable to find a
near optimal flow assignment for large-scale Data Center networks. The suggested
approach of traffic merging can reduce energy consumption of active switches. With
verylightload,thiskindoftrafficmergingcansave20–40%energycostcomparedto
thewell-establishedelastic treeapproach.Kuehnintroduced anovelmethodtoreduce
task graphs with generally distributed task processing times to a single virtual job
processingtime.Lookingataverydifferentproblem,i.e.,thechallengeofdealingwith
1 FP7STREP#609304,www.dc4cities.eu.
VI Preface
frequent blackouts from an unstable power grid, Al-Salim et al. proposed a cyclic
blackout mitigation through shifting of HVAC loads by means of queuing optimiza-
tion. Finally, in a work by Postema and Haverkort a set of stochastic petri net models
wasappliedtotheanalysisoftrade-offsbetweenperformanceandpowerconsumption
of Data Center. This modeling approach is meant to support decisions in the early
design stage of a Data Center.
Thesecondpartoftheproceedingscontainsfourpapersfocusedonthefutureroleof
Data Centers in Europe. In this session Anghel et al. presented the European project
GEYSER.ThisprojectisaimedatintegratingDataCentersintoSmartGridsandSmart
Cities and its scope is to realize an optimized intelligent pervasive sensing and mon-
itoring infrastructure. Gribaudo et al. in their paper proposed an analysis of the
influence of application deployment on energy consumption based on the European
projectECO2Clouds.Theauthorsinvestigateddifferent waystodeployanapplication
in clouds and analyzed simultaneously energy consumption and system performances
for each deployment configuration. In their paper “Minimization of Costs and Energy
ConsumptioninDataCentersbyaWorkload BasedCapacityManagement”DaCosta
et al. proposed a holistic view to Data Center modeling including workloads and
cooling.TheyintroduceddynamicpowercappingtoDataCenterenergymanagement.
Dupont’sobjectiveforDataCenterenergymanagementwastogiveacontributionfor
making Data Centers more energy aware with regard to the availability of renewable
energy.Tothispurposeanenergy-awarevirtualmachinesmanagerbasedonConstraint
Programming (Plug4Green) was applied.
The third part of the proceedings discusses energy efficiency metrics for Data
Centers.Capozzolietal.presentedacritical reviewofperformancemetrics forenergy
efficiency in Data Centers aiming to demonstrate the crucial role of thermal manage-
mentforenergysaving.Schlittetal.intheirpapersuggestednewmetricsbeyondPUE,
capable to consider the adaptability of infrastructure to IT power: the infrastructure
poweradaptability(IPA)metricrepresentingthepoweradaptabilityoftheDataCenter
infrastructure in combination with the power variability (PVar).
The workshop also included three additional presentations: an introduction to the
EU projects All4Green by Sonja Klingert (University of Mannheim) and DC4Cities
by Marta Chinnici (ENEA, Italian National Agency for New Technologies, Energy
and Sustainable Economic Development, Italy), as well as a keynote speech by
Ian F. Bitterlin (CTO Emerson Network Power Systems, Visiting Professor at Uni-
versityofLeeds)abouttheproblemofmushroomingdatagrowthwhichisspurringthe
energy growth of the global Data Center industry and can only partly be offset by
technical evolution and innovation.
WewouldliketothankIanF.Bitterlinandallauthorsfortheircontributionstothe
thirdvolumeoftheE2DCproceeding,andalsothereviewersfortheireffort:theyboth
helped in selecting the best papers and improving the initial submissions. Also, thank
you to the Session Chairs Hermann de Meer (University of Passau), Jaume Salom
(IREC), and Alfonso Capozzoli (Politecnico di Torino). And we are grateful for an
Preface VII
interested and interesting audience, who with lively discussions helped in turning the
workshop yet again into a successful event.
Finally, we are grateful for the strong support from the European Commission and
the ICT FP7 All4Green project.
October 2014 Sonja Klingert
Marta Chinnici
Milagros Rey Porto
Organization
Workshop Chairs
Sonja Klingert University of Mannheim, Germany
Marta Chinnici ENEA, Italy
Milagros Rey Porto Gas Natural Fenosa, Spain
Technical Program Committee
Hermann de Meer University of Passau, Germany
Daniel Gmach HPLabs, USA
Jorjeta Jetcheva Fujitsu Laboratories of America, USA
Paul Kuehn University of Stuttgart, Germany
Eric Madeleine Inria, France
Maria Perez Ortega GFI, Spain
Mary Ann Piette Lawrence Berkeley National Laboratory, USA
Barbara Pernici Politecnico di Milano, Italy
Gunnar Schomaker OFFIS, Germany
Shaolei Ren Florida International University, USA
Tomasz Siewierski Technical University of Lodz
Sponsoring Institutions
EU FP7 Project DC4Cities (#609304)
University of Mannheim, Germany
ENEA, Italy
Gas Natural Fenosa, Spain
Contents
Energy Optimization Algorithms and Models
Agile Traffic Merging for DCNs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Qing Yi and Suresh Singh
Performance and Energy Efficiency of Parallel Processing in Data
Center Environments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Paul J. Kuehn
Cyclic Blackout Mitigation Through HVAC Shifted Queue Optimization. . . . 34
Kasim Al-Salim, Ivan Andonovic, and Craig Michie
Stochastic Petri Net Models for the Analysis of Trade-Offs in Data Centres
with Power Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Björn F. Postema and Boudewijn R. Haverkort
The Future Role of Data Centres in Europe
GEYSER: Enabling Green Data Centres in Smart Cities . . . . . . . . . . . . . . . 71
Ionut Anghel, Massimo Bertoncini, Tudor Cioara, Marco Cupelli,
Vasiliki Georgiadou, Pooyan Jahangiri, Antonello Monti, Seán Murphy,
Anthony Schoofs, and Terpsi Velivassaki
Analysis of the Influence of Application Deployment
on Energy Consumption. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Marco Gribaudo, Thi Thao Nguyen Ho, Barbara Pernici,
and Giuseppe Serazzi
Minimization of Costs and Energy Consumption in a Data Center
by a Workload-Based Capacity Management . . . . . . . . . . . . . . . . . . . . . . . 102
Georges Da Costa, Ariel Oleksiak, Wojciech Piatek, Jaume Salom,
and Laura Sisó
Building Application Profiles to Allow a Better Usage of the Renewable
Energies in Data Centres . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
Corentin Dupont
Energy Efficiency Metrics for Data Centres
Review on Performance Metrics for Energy Efficiency in Data Center:
The Role of Thermal Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
Alfonso Capozzoli, Marta Chinnici, Marco Perino, and Gianluca Serale