Emanouil Atanassov Grid Technologies and Applications Institute for Parallel Processsing Bulgarian Academy of Sciences R–ECFA meeting, Sofia, 7-8 May 2010 SEERA-EI
Jan 14, 2016
Emanouil AtanassovGrid Technologies and Applications Institute for Parallel Processsing Bulgarian Academy of Sciences
R–ECFA meeting, Sofia, 7-8 May 2010
SEERA-EI
Overview
SEERA-EI PSC02 meeting, Bucharest, 6-7 July 2009 2
• Motivation• EGEE project• The regional dimension• Grid computing resources in Bulgaria• Grid applications• Perspectives
Motivation
R–ECFA meeting, 7-8 May 2010, Sofia 3
Grid – coordinated resource sharing and problem solving indynamic, multi-institutional virtual organizations
GEANT Network infrastructure, National Grid
resource centers
Su
pp
ort
, tra
inin
g,
ed
uc
ati
on
Collaborations
Грид
R–ECFA meeting, 7-8 May 2010, Sofia 4
The Bulgarian Grid consortium
Founders: Institute for Parallel Processing (IPP-BAS, former CLPP-BAS) and Institute for Nuclear Research and Nuclear Energy (INRNE-BAS) – (Agreement signed in September 2002)
Joined by: Institute of Mathematics and Informatics (IMI), Bulg. Academy of Sciences Institute of Mechanics (IM), Bulgarian Academy of Sciences Faculty of Mathematics and Informatics (FMI), Plovdiv University Computer Science Department (CSD), American University in Bulgaria
(AUBG) Department of Information Technologies at the Faculty of Mathematics
and Informatics (DIT-FMI), Sofia University. Institute of Electrochemistry and Energy Systems (IEES), BAS. Institute of Astronomy, BAS University of Mining and Geology Geophysical Institute ( GPhI), BAS Faculty of Physics – Sofia University Institute of Organic Chemistry with centre of Phytochemistry University of National and World Economics
Main purposes: Sharing of resources and expertise in Grid technologies Development of Grid applications
R–ECFA meeting, 7-8 May 2010, Sofia 5
EGEE project
Services Continuous monitoring of grid services Support for diverse research disciplinesArcheology, Astronomy,Astrophysics,Civil Protection,Comp. Chemistry,Earth SciencesFinance,Fusion,Geophysics,High Energy Physics, Life Sciences,Multimedia,Material Sciences, ...MiddlewareProduction middleware distributed under business friendly open source licence
Implements a service-oriented architecture that virtualises resources
Adheres to recommendations on web service inter-operability and evolving towards emerging standards
User Support - Managed process from first contact through to production usageTrainingExpertise in grid-enabling applicationsOnline helpdeskNetworking events
More than 100 000 CPU, 10 PB disk storageMoving from regional to national model
R–ECFA meeting, 7-8 May 2010, Sofia 6
Regional dimension
GRNET.gr
MTA SZTAKI .hu
IPP .bg
ICI .ro
TUBITAK.tr
UPT.al
UKIM.mk
UoBL.ba
RBI.hr
UOB.rs
RENAM.md
UOM.me
GRENA.ge
IIAP-NAS-RA.am
SEEFIRE + Geant Support
SEE-GRID
SeismologyVO
MeteorologyVO
EnvironmentalVO
SeismologyVO
MeteorologyVO
EnvironmentalVO
SEE-SCI
e-Infrastructure
The regional Grid infrastructure is now fully integrated into the European Grid infrastructure. In the SEE-GRID-SCI project it serves as the basis for enabling new Grid applications from the domains of seismology, meteorology, environmental modelling and environmental protection SEEREN initiative has established regional network and SEE-LIGHT project is currently establishing dark-fibre backbone in the region. Bulgaria now has a GEANT Point of Presence at IPP-BAS. SEERA-EI will capitalise on this momentum and link national-level programme managers and provide an open forum for information exchange, in order to enable coordination of national programmes in eInfrastructures, and to set the framework for a common regional agenda.
Grid computing resources in Bulgaria
Sites CPU Storage
BG01-IPP 636 1 TB (48TB)
BG02-IM 21 0.01 TB
BG-INRNE 80 1 TB
BG03-NGCC 200 8 TB
BG04-ACAD 80 2 TB
BG05-SUGrid 28 0.2 TB
BG06-GPHI 44 1 TB (15TB)
BG08-MADARA 800 24 TB(108TB)
Total 1853 37 TB
•BG03-NGCC was installed at IPP-BAS and delivered more than 50% of CMS usage in Bulgaria in the period 2009-2010. •The biggest Bulgarian Grid clusters have low-latency DDR Infiniband interconnect and achieve more than 90% efficiency in the LINPACK Benchmark (>3 TF each), which makes them especially suitable for parallel (MPI) applications from computational chemistry, environmental modelling, etc.•Thanks to the two NSFBG infrastructure projects Bulgaria became a leader in CPU resources in South Eastern Europe
R–ECFA meeting, 7-8 May 2010, Sofia 8
Supported Virtual Organizations
Bugarian Grid sites support the major HEP VOs – CMS, Alice, Atlas, LHCb, na48, geant4Other EGEE VOs supported:ESR, Biomed, MAGICRegional VOs: seegrid, see, env.see-gri-sci.eu, meteo.see-grid-sci.eu, seismo.see-grid-sci.euNational: mm-comp-chem.grid.acad.bgbgedu.grid.acad.bg, …
R–ECFA meeting, 7-8 May 2010, Sofia 9
SALUTE - Stochastic ALgorithms for Ultra-fast Transport in sEmiconductors
SALUTE is a grid application, which integrates a set of novel Monte Carlo and quasi-Monte Carlo algorithms for solving computationally intensive problems from ultrafast carrier transport in semiconductors.
SALUTE studies memory and quantum effects during the relaxation process due to electron-phonon interaction in semiconductors; present version explores electron kinetics in GaAs nano-wires.
R–ECFA meeting, 7-8 May 2010, Sofia 10
Bulgarian Grid Applications - Chemistry
BG08-MADARA is used exclusively for modeling and simulations in the field of chemistry and materials scienceMore than 1 million CPU hours delivered. Most popular job type – 16 CPUs for 29 hours on average, 662959 hours total, next one in popularity – 32 CPUs for 37 hours on average, 179772 hours total
R–ECFA meeting, 7-8 May 2010, Sofia 11
Multiscale Atmospheric Composition Modeling
• GRID environment based integrated, multi-scale Balkan region oriented AQ modelling system• Study the atmospheric pollution transport and transformation processes (accounting also for heterogeneous chemistry and the importance of aerosols for air quality and climate) from urban to local to regional (Balkan) scales; • Track and characterize the main pathways and processes that lead to atmospheric composition formation in different scales; • Provide high quality scientifically robust assessments of the air quality and its origin, thus facilitating formulation of pollution mitigation strategies at national and Balkan level.
4 0 5 0 6 0 7 0 8 0 9 0
Ju lian day
0
4 0
8 0
1 2 0
1 6 0
PM
10 [u
g/m
^^3]
D E02simulated-30kmm easuredsimulated-10 km
4 0 5 0 6 0 7 0 8 0 9 0
Ju lian day
0
4 0
8 0
1 2 0
1 6 0
PM
10 [
ug/m
^^3]
DE03sim ulated-30kmm easuredsim ulated-10km
4 0 5 0 6 0 7 0 8 0 9 0
Ju lian day
0
4 0
8 0
1 2 0
1 6 0
PM
10 [
ug/m
^^3]
DE04simulated-30kmmeasuredsimulated-10km
4 0 5 0 6 0 7 0 8 0 9 0
Ju lian day
0
4 0
8 0
1 2 0
1 6 0
PM
10 [
ug/m
^^3]
DE05simulated-30kmm easuredsimulated-10km
4 0 5 0 6 0 7 0 8 0 9 0
Ju lian day
0
4 0
8 0
1 2 0
1 6 0
PM
10 [
ug/m
^^3]
DE07sim ulated-30kmm easuredsim ulated-10km
4 0 5 0 6 0 7 0 8 0 9 0
Ju lian day
0
4 0
8 0
1 2 0
1 6 0
PM
10 [
ug/m
^^3]
DE08simulated-30kmmeasuredsimulated-10km
R–ECFA meeting, 7-8 May 2010, Sofia 12
MSACM Grid-mode tests: Country-to-country “blame” matrixes on the Balkans according to EMEP methodology
Blame matrix for oxidized sulphur total deposition, May-July 2000, 1000 t(S).
Emitter Receiver
BG GR RO other all sources
BG 14.772 0.381 3.021 5.838 31.717 GR 3.612 4.230 0.795 6.281 9.058 RO 3.278 0.200 18.203 16.652 28.251 other 10.055 4.246 6.232 64.605 93.376 deposited 31.717 9.058 28.251 93.376 162.401 total emission [S] 72.425 6.708 68.374 180.789 348.297
% Rec/Emit 43.792 33.913 41.318 51.649 46.627
Blame matrix for oxidized nitrogen total deposition, May-July 2000, 1000 t(N).
Emitter
Receiver BG GR RO other all sources
BG 1.065 0.011 0.137 0.150 1.375 GR 0.077 0.890 0.000 0.458 1.015 RO 0.108 0.000 1.916 0.481 2.202 other 0.125 0.114 0.149 4.496 5.584 deposited 1.375 1.015 2.202 5.584 10.176 total emission [N] 18.317 20.406 44.210 91.430 174.363
% Rec/Emit 7.506 4.976 4.981 6.107 5.836
Blame matrix for reduced nitrogen total deposition, May-July 2000, 1000 t(N).
Emitter
Receiver BG GR RO other all
sources BG 3.178 0.299 1.232 1.974 5.598 GR 0.455 3.004 0.297 2.119 5.058 RO 0.682 0.090 11.843 5.156 16.403 other 1.282 1.664 3.031 23.874 33.123 deposited 5.598 5.058 16.403 33.123 60.182 total emission [N] 18.317 20.406 44.210 91.430 174.363
% Rec/Emit 30.561 24.785 37.103 36.228 34.515
Averaged for a “typical” day in July fields of roadc [%] for ozone, PM2.5, nitrogen
compounds, hydrocarbons, CAR_PHE, PMcoarse, organic Nitrates and SO2
R–ECFA meeting, 7-8 May 2010, Sofia 13
MSERRHSA: modelling accidental Chlorine release in “VEREJA-HIM” JAMBOL – averaged surface HCl concentrations for January 2008 for
two release times
R–ECFA meeting, 7-8 May 2010, Sofia 14
Conclusion
From the Communication from European Commission “ICT Infrastructures for e-Science” „Research in 2020 cannot be imagined without the intensive use of sophisticatede-Infrastructures, so Europe needs to commit to a renewed strategy to tackle theassociated challenges and priorities.”„Striving for world leadership in e-Science, establishing e-Infrastructures as asustainable utility and exploiting them to promote innovation are the three vectorsof a renewed European strategy to support the ground-breaking science of 2020 andbeyond.”