author - title- date - n° 1 high energy phisics and grids lnf- 9 may 2002 federico ruggieri –...
TRANSCRIPT
Author - Title- Date - n° 1
High Energy Phisics and GRIDs
LNF- 9 May 2002
Federico Ruggieri – INFN CNAF - Bologna
E-Mail: [email protected]
Author - Title- Date - n° 2
To learn more
DataGrid: www.eu-datagrid.org
Globus: www.globus.org
VI F.P.: www.cordis.lu/fp6
INFN-GRID: www.infn.it/grid
LHC Computing GRID: lhcgrid.web.cern.ch
Author - Title- Date - n° 3
Summary
The GRID concepts and objectives
Why HEP needs/promotes GRIDs
DataGRID program, architecture and achievements
INFN GRID and DataGRID Testbeds
Other GRID Projects
US GRIDs and coordination issues.
6th Framework Program
Author - Title- Date - n° 4
GRID concepts and objectives
Extend the Web concept of Accessibilty to all kind of resources.
Make it “invisible” to the end user that doesen’t care of the real topology and configuration.
Use as much as possible of what is already widely deployed: Network: Internet and TCP/IP Protocols: http, TCP, UDP, …. Operating Systems: Linux, Solaris, ….. Batch Systems: PBS, LSF, Condor, ….. Storage: Disks, HPSS, HSM, CASTOR, ….. Directory Services: LDAP, …. Certificates: X509
Create a Middle-ware layer between the facilities/services and the applications
Author - Title- Date - n° 5
GRID an extention of the WEB concept
On-demand creation of powerfulOn-demand creation of powerful virtual computing and data virtual computing and data systemssystems
Grid: Flexible and High Performance Access to All kinds of Resources.
Sensor nets
http://
http://
Web: Uniform Access to the Informations
Data Stores
Computers
Softwarecatalogs
Colleagues
0
50
100
150
200
250
300A
ust
ria
Cze
ch R
ep
ub
lic
De
nm
ark
Fin
lan
d
Fra
nce
Ge
rma
ny
Gre
ece
Ita
ly
Ne
the
rlan
ds
No
rwa
y
Po
lan
d
Po
rtu
ga
l
Slo
vac
Re
pu
blic
Sp
ain
Sw
ed
en
Sw
itze
rlan
d
Un
ited
Kin
gd
om
CE
RN
Arm
en
ia
Au
stra
lia
Aze
rba
ijan
Re
pu
blic
Re
pu
blic
of
Be
laru
s
Bra
zil
Ca
na
da
Ch
ina
PR
Re
pu
blic
of
Ge
org
ia
Isra
el
Jap
an
Mo
rocc
o
Ro
ma
nia
Ru
ssia
JIN
R D
ub
na
Slo
ven
ia
Taiw
an
Tu
rke
y
Un
ited
Sta
tes
Author - Title- Date - n° 6
GRID concepts and objectives
General approach for the distribution of the Electric PowerSystem of Generators, High Voltage, Transformers, Distribution Network that bring Electric Power to every user
From T. Priol 4th Datagrid Conference Paris
Author - Title- Date - n° 7
The Computing Grid Concept
From T. Priol 4th Datagrid Conference Paris
General approach for the distribution of the Computing PowerHigh Bandwidth Internet Network provide access from home PC’s to “Infinite” Computing, Storage and Application ResourcesYou pay what you use
Author - Title- Date - n° 8
The Globus Project
“Dependable, consistent, pervasive access to [high-end] resources”
• Dependable: Can provide performance and functionality guarantees
• Consistent: Uniform interfaces to a wide variety of resources
• Pervasive: Ability to “plug in” from anywhere
Author - Title- Date - n° 9
Layered Structure
Applications
Core ServicesMetacomputing
Directory Service
GRAMGlobus
Security Interface
Heartbeat Monitor
Nexus
Gloperf
Local Services
LSF
Condor MPI
NQEEasy
TCP
SolarisIrixAIX
UDP
High-level Services and Tools
DUROC globusrunMPI Nimrod/GMPI-IO CC++
GlobusView Testbed Status
GASS
Author - Title- Date - n° 10
Why HEP needs/promotes GRIDs
LHC Computing requires big amounts of distributed resources: CERN is foreseen to provide a maximum of 30% of these resources
Several thousands of researches, all around the world, want to produce MC and Analyse Data with as less restriction as possible.
10-20 years machine lifetime need a robust and scalable distributed system.
Many present experiments (BaBar, CDF, D0, VIRGO, ...) seem to need the same functionalities right now (and they already developed some of them).
The solution has to live many years and should, possibly, not be HEP specific:
Long term support is much harder if you are the only customer
New Ideas can come from other applications and/or developers
Author - Title- Date - n° 11
A very large distributed comunity
CMS: 1800 physicists150 institutes32 countries
Just as an example
Author - Title- Date - n° 12
MONARC Architecture
Tier 1R.C.
Tier 1R.C.
Tier 1R.C.
Net
wor
k
Tier 2
Tier 2
Tier 2
Tier 2
Tier 2
Tier 2
Net
wor
k
CERN(Tier 0)
Net
wor
kN
etw
ork
Author - Title- Date - n° 13
Foreseen required resources
CERN (Sum of all the experiments): Mass Storage: 10 Peta Bytes (1015 B)/yr
disk: 2 PB
CPU: 2 MSI95 (PC today ~ 30-40SI95) = 20 MSPECint2000
For each Multi-Experiment Tier 1 : Mass Storage: 3 PB/yr
disk: 1.5 PB
CPU: 1 M SI95
Networking Tier 0 --> Tier 1: 2 Gbps
Author - Title- Date - n° 14
CERN Centre Physics Computing Capacity
0
1
2
3
4
5
6
1988 1989 1990 1991 1992 1993 1994year
Th
ou
sa
nd
s o
f C
ER
N U
nit
s
IBM 9000 mainframe decomissioned
Moore's law (based on 1988)
SHIFT Project at CERN
CERN Centre Physics Computing Capacity
0
20
40
60
80
100
120
1988 1990 1992 1994 1996 1998 2000year
Th
ou
san
ds
of
CE
RN
Un
its
Mainframes decomissioned
First PC services
Moore's law(based on 1988)
CERN RD47 project
RISC decomissioning agreed
LEP
LHC
Evolution of Computing Capacity at CERN
Moore’s law –some measure of the capacity technology advances provide for a constant number of processors or investment (based on 2000)
Jan 2000:3.5K SI95
Author - Title- Date - n° 16
DataGRID the Flagship Project in EU
Contract Started officially on the 1st of January 2001 for approx. 9.8 M€ in 3 years.
First EU Review succesfully passed on the 1st of March 2002.
Successful deployment of the Testbed 1 (> 140K code lines in 10 languages: C, C++, Java, Perl, etc.).
Handful of main sites in Europe (CERN, CNAF, LYON, NIKHEF, RAL).
Many other sites in Italy: MI, BO, TO, PD (+LNL), CT, PI, ....
A production testbed will soon be available.
Testbed 2 is foreseen in fall (september-october) this year.
Author - Title- Date - n° 17
DataGRID Project
European level coordination of national initiatives & projects.
Main goals:– Middleware for fabric & Grid management– Large scale testbed - major fraction of one LHC experiment– Production quality HEP demonstrations (real users, real
applications, real data)– Other science demonstrations– Three years phased developments & demos
Complementary to other GRID projects
Synergy with other activities (GRID Forum, Industry and Research Forum)
Author - Title- Date - n° 18
Participants
Main partners: CERN, INFN(I), CNRS(F), PPARC(UK), NIKHEF(NL), ESA-Earth Observation
Other sciences: KNMI(NL), Biology, Medicine
Industrial participation: CS SI/F, DataMat/I, IBM/UK
Associated partners: Czech Republic, Finland, Germany, Hungary, Spain, Sweden (mostly computer scientists)
Formal collaboration with USA being established
Industry and Research Project Forum with representatives from:– Denmark, Greece, Israel, Japan, Norway, Poland, Portugal, Russia,
Switzerland, etc.
Author - Title- Date - n° 19
Workpackages
INFN
WP 1 Grid Workload Management (F. Preltz & M. Sgaravatto/INFN)
WP 2 Grid Data Management (P. Kunstz/CERN) WP 3 Grid Monitoring services (S. Fisher/PPARC) WP 4 Fabric Management (O. Barring/CERN) WP 5 Mass Storage Management (J. Gordon/PPARC) WP 6 Integration Testbed (F. Etienne/CNRS) WP 7 Network Services (P. Primet/INRIA-CNRS) WP 8 HEP Applications (F. Carminati/CERN) WP 9 EO Science Applications (L. Fusco/ESA) WP 10 Biology Applications (V. Breton/CNRS) WP 11 Dissemination (M. Lancia/CNR) WP 12 Project Management (F. Gagliardi/CERN)
Author - Title- Date - n° 20
WP 1 GRID Workload Management
Goal: define and implement a suitable architecture for distributed scheduling and resource management in a GRID environment.
Issues: Optimal co-allocation of data, CPU and network for specific
“grid/network-aware” jobs
Distributed scheduling (data and/or code migration) of unscheduled/scheduled jobs
Uniform interface to various local resource managers
Priorities, policies on resource (CPU, Data, Network) usage
Author - Title- Date - n° 21
DataGrid Architecture
Collective ServicesCollective Services
Information &
Monitoring
Information &
Monitoring
Replica ManagerReplica
ManagerGrid
SchedulerGrid
Scheduler
Local ApplicationLocal Application Local DatabaseLocal Database
Underlying Grid ServicesUnderlying Grid Services
Computing Element Services
Computing Element Services
Authorization Authentication and Accounting
Authorization Authentication and Accounting
Replica CatalogReplica Catalog
Storage Element Services
Storage Element Services
SQL Database Services
SQL Database Services
Fabric servicesFabric services
ConfigurationManagement
ConfigurationManagement
Node Installation &Management
Node Installation &Management
Monitoringand
Fault Tolerance
Monitoringand
Fault Tolerance
Resource Management
Resource Management
Fabric StorageManagement
Fabric StorageManagement
Grid
Fabric
Local Computing
Grid Grid Application LayerGrid Application Layer
Data Management
Data Management
Job Management
Job Management
Metadata Management
Metadata Management
Object to File Mapping
Object to File Mapping
Service Index
Service Index
Author - Title- Date - n° 22
A Job Submission Example
UIJDL
Logging &Logging &Book-keepingBook-keeping
ResourceResourceBrokerBroker
Output “sandbox”
Input “sandbox”
Job SubmissionJob SubmissionServiceService
StorageStorageElementElement
ComputeComputeElementElement
Brokerinfo
Output “sandbox”
Input “sandbox”
Information Information ServiceService
Job Status
ReplicaReplicaCatalogueCatalogue
Author.&Authen. Job
Subm
it
Job Q
uery
Job Status
Author - Title- Date - n° 23
INFN GRID Evolution Path
ATLAS
INTEGRATIONQuantumGRID
TESTBED 1
GRIDCMS ALICE
ATLASLHCb
VIRGO
TIME
CO
MP
LE
XIT
Y CMS
ALICELHCb
VIRGO
APEARGOBaBar
CDF
Author - Title- Date - n° 24
INFN GRID & DataGRID Testbeds
DataGrid Appl. & partners VOs
INFN-GridVO
2 grid domains:• 6 sites for DataGrid Testbed1 • 6+12sites for national Testbed1
CERN
Lyon/in2p3
RAL/pparc
nikhefmi to
ba
ca
nale pd/lnl
Cnaf/boctpi
roma
ts
ge
pvrm3
fepr
Each InfnGrid site can join the DataGrid Testbed if needed
Author - Title- Date - n° 25
Other GRID Projects (not exaustive)
EU Funded: DataTAG
CrossGRID
EuroGRID
........
National: INFN-GRID (IT), GRIDPP (UK), France, Netherlands, .......
United States PPDG (DOE) :
GRIPHYN (NSF) :
IVDGL (NSF):
+ HEP Specific CERN LHC Computing GRID Project
Author - Title- Date - n° 27
Catania
Bologna
Padova/LNL
Torino
Cagliari
Roma
Milano
To USA
To Russia/Japan
The prototype INFN DataGrid testbed
Cern
Author - Title- Date - n° 28
NLNLSURFnet
CERN
UKUKSuperJANET4
AbileneAbilene
ESNETESNET
MRENMREN
ITITGARR-B
GEANT
NewYork
STAR-TAP
STAR-LIGHT
DataTAG project
Two main focus:
Grid applied network research; 2.5 Gbps lambda with Star-Light for network research
Interoperability between Grids in EU and US (Managed by INFN : 0.6 M€)
US partnership: iVDGL project (10M$)
Main PartnersCERN, INFN,UvA(NL) PPARC(UK), INRIA(FR)
Author - Title- Date - n° 29
EU VI Framework Program
Integrated Project Proposal
Letter of Intent within 15 May 2002
350 M€ Budget Line (GRID + GEANT)
Infrastructure deployment for e-science and research.
E-governament and e-business tests and prototypes
High school access
US and Trans-continental interoperability
Key role of Industry.
Author - Title- Date - n° 30
VI Framework Program GRID Initiative
ConsultingPrototypingDeployment
ConsultingTraining Courses
DisseminationForum
Tools and Service Development
Creation and support E- Science centres
Applications inOther Sciences
EI ROforum
S/W HardeningGLOBUS
EuroGrid, Gridlab etc.
Semantic GRI DDatabaseSecurity
European I nfrastructure
Integrated ProjectENABLI NG GRI DS ESCI ENCE EUROPE
EGEE
National eScienceCentres I ndustry
Applications
SMEs developingGrid-enabled Applications
I ndustry Outreach
ModulableTestbeds
R&D Agenda
Deployment withI T I ndustry
ScienceOutreach
INFN proposal to other Italian research organizations, governmentand Industries to develop a common EU GeS initiative (~150 M€)involving CERN and most of eScience EU activities and ICTIndustries
Area’s ofactivity