1. The LHC Computing Grid Project Status & Plans
LCG
CERN-US Co-operation Committee
18 June 2004
Jürgen Knobloch
IT Department, CERN
This file is available at:

2. The LHC Computing Grid Project - LCG
Collaboration
LHC Experiments
Grid projects: Europe, US
Regional & national centres worldwide
Choices
Adopt Grid technology.
Go for a “Tier” hierarchy.
Use Intel CPUs in standard PCs
Use LINUX operating system.
Goal
Prepare and deploy the computing environment for the analysis of data from the LHC detectors.
Lab m
Uni x
grid for a
regional group
Uni a
CERN Tier 1
Lab a UK
USA FNAL
Tier-1
Tier3
physics
department
USA BNL
France
Uni n
Tier-2
CERN Tier 0
Japan
Italy 
Lab b
Germany
Lab c
grid for a physics
study group 
Uni y 
Uni b
Desktop

3. Operational Management of the Project
Applications Area
Development environment
Joint projects
Data management
Distributed analysis
Middleware Area – now EGEE
Provision of a base set of grid middleware – acquisition, development, integration, testing, support
ARDA
A Realisation of Distributed Analysis for LHC
CERN Fabric Area
Large cluster management
Data recording
Cluster technology
Networking
Computing service at CERN
Grid Deployment Area
Establishing and managing the Grid Service - Middleware certification, security, operations, registration, authorisation, accounting
Joint with a new European project EGEE
Phase 1 – development of common software prototyping and operation of a pilot computing service
Phase 2 – acquire, build and operate the LHC computing service
Enabling Grids for e-Science in Europe

4. Sites in LCG-2/EGEE-0: June 18 2004
Austria
U-Innsbruck
Canada
Triumf
Alberta
Carleton
Montreal
Toronto
Czech Republic
Prague-FZU
Prague-CESNET
France
CC-IN2P3
Clermont-Ferrand
Germany
FZK
Aachen
DESY
GSI
Karlsruhe
Wuppertal
Greece
HellasGrid
Hungary
Budapest
India
TIFR
Israel
Tel-Aviv
Weizmann
Italy
CNAF
Frascati
Legnaro
Milano
Napoli
Roma
Torino
Japan
Tokyo
Netherlands
NIKHEF
SARA
Pakistan
NCP
Poland
Krakow
Portugal
LIP
Russia
SINP-Moscow
JINR-Dubna
Spain
PIC
UAM
USC
UB-Barcelona
IFCA
CIEMAT
IFIC
Switzerland
CERN
CSCS
Taiwan
ASCC
IPAS
NCU UK
RAL
Birmingham
Cavendish
Glasgow
Imperial
Lancaster
Manchester
QMUL
RAL-PP
Sheffield
UCL US
BNL
FNAL HP
Puerto-Rico
22 Countries
62 Sites (48 Europe, 2 US, 5 Canada, 6 Asia, 1 HP)
Coming: New Zealand, China,
other HP (Brazil, Singapore)
4000 CPUs

5. LCG-2 sites – status 17 June 2004

6. LCG Service Status Certification and distribution process established
Middleware package – components from –
European DataGrid (EDG)
US (Globus, Condor, PPDG, GriPhyN)  the Virtual Data Toolkit
Principles for registration and security agreed
Grid Operations Centre at Rutherford Lab (UK)
A second centre is coming online at Academia Sinica in Taipei
Call Centre FZK (Karlsruhe)
LCG-2 software released February 2004
62 centres connected with more than 4000 processors
Four collaborations run data challenges on the grid
Status on 24 July 2003 – I will an update on the 11th August
There are two activities going on in parallel:
-- Pre-production software is being deployed to 10 Regional Centres (Academia Sinica Taipei, BNL, CERN, CNAF, FNAL, FZK, IN2P3 Lyon, Moscow State Univ., RAL, Univ. Tokyo). This is a final validation of the distribution and installation process and tools, and establishes the network of service support people, but the Grid middleware is not the final version yet. About half of the sites have successfully installed this – a few (including BNL and Lyon) are very slow.
-- Final testing is under way of the first production set of Grid middleware. This is being done by a group of grid experts from the LHC experiments (the “loose cannons”). There are a few “show stoppers” and about 15 very serious bugs.
The target is to install the production software by the end of the month – one week from now!

7. Data challenges Now Summer ‘04 Fall ‘04
The 4 LHC experiments currently run data challenges using the LHC computing grid
Part 1: World-wide production of simulated data
Job submission, resource allocation and monitoring
Catalog of distributed data
Part 2: Test of Tier-0 operation
Continuous (24 x 7) recording of data up to 450 MB/s per experiment (target for ALICE in 2005: 750 MB/s)
First pass data reconstruction and analysis
Distribute data in real-time to Tier-1 centres
Part 3: Distributed analysis on the Grid
Access of data from anywhere in the world – in an organized as well as in a “chaotic” access pattern
Now
Summer ‘04
Fall ‘04

8. Grid operation
After 3 months of intensive use the basic middleware of LCG-2 is proving to be reliable – although there are many issues of functionality, scaling, performance, scheduling
The grid deployment process is working well –
Integration – certification – debugging
Installation and validation of new sites
Constructive and useful feedback from first data challenges – especially on data issues from CMS
Interoperability with Grid3 in the US being studied (presentation by FNAL at Grid Deployment Board in May)
Implementation of a common interface to Mass Storage completed or in plan at all Tier-1 Centres
Proposal for convergence on a single Linux flavour – based on the FNAL Scientific Linux package

9. Service Challenges Exercise the operations and support infrastructure
Gain experience in service management
Uncover problems with long-term operation
Focus on
Data management, batch production and analysis
Reliable data transfer
Integration of high bandwidth networking
Also: massive job submission, test reaction to security incidents
Target by end 2004 –
Robust and reliable data management services in continuous operation between CERN, Tier-1 and large Tier-2 centres
Sufficient experience with sustained high performance data transfer to guide wide area network planning
The Service Challenges are a complement to the experiment Data Challenges

10. Networking 5.44 Gb/s 1.1 TB in 30 min. 6.25 Gb/s 20 April 04
Key element for LHC computing strategy
Aiming at sustained 500 MB/s by end 2004 – requiring 10 Gb/s networks to some Tier-1 centres based on existing facilities
5.44 Gb/s
1.1 TB in 30 min.
6.25 Gb/s
20 April 04

11. LCG-2 and Next Generation Middleware
prototyping
product
2004
2005
LCG-3
LCG-2 focus on production, large-scale data handling
The service for the 2004 data challenges
Provides experience on operating and managing a global grid service
Strong development programme driven by data challenge experience
Evolves to LCG-3 as components progressively replaced with new middleware
Next generation middleware focus on analysis
Developed by EGEE project in collaboration with VDT (US)
LHC applications and users closely involved in prototyping & development (ARDA project)
Short development cycles
Completed components integrated in LCG-2

12. US participation - some examples
US LHC funds Miron Livny on ARDA/EGEE middleware
US CMS working on end to end ARDA prototypes for the fall.
US CMS will participate in CCS milestone for analysis across grids.
NSF funds VDT/EGEE joint effort.
Wisconsin certification testbeds for national NMI testing, VDT, LCG and EGEE middleware... etc. encourage coherence across the middleware versions.
PPDG helps fund ATLAS ARDA representative.
PPDG extension proposal is funded. Enables some immediate attention to Open Science Grid needs.
Open Science Grid “Technical Groups” scope explicitly includes cooperation with the EGEE/LCG peers groups.
A single Linux flavour – based on the FNAL Scientific Linux package

13. Phase 2 Planning Outline
June Establish editorial board for LCG TDR
September 2004 – Consolidated Tier-1, Tier-2 Regional Centre Plan
Background for the draft MoU  October C-RRB
Revised version of basic computing models
Revised estimates of overall Tier-1, Tier-2 resources
Current state of commitments of resources Regional Centres ↔ Experiments
High-level plan for ramping up the Tier-1 and large Tier-2 centres
October 2004 C-RRB – Draft MoU
End Initial computing models agreed by experiments
April 2005 C-RRB - Final MoU
End June TDR

14. Planning groups High level planning group – chair Les Robertson
Overall picture & implied resources at centres
Representatives from the experiments and LCG management
Representatives from three regional centres – Italy, Spain, US
Grid Deployment Area steering group – chair Ian Bird
Interoperability, resources, overview Service Challenges
Representatives from regional centres
Service challenge group – chair Bernd Panzer
Carrying out service challenges
People working on service challenges
Networking – chair David Foster
Working towards required end-to-end bandwidth
MoU taskforce – chair David Jacobs
Draft LCG MoU and one for each of the four experiments
Representatives from experiments, six countries, LCG management
TDR editorial board – chair Jürgen Knobloch

15. 102 FTE-years missing for Phase 2
LCG Staffing for the Common Activities at CERN
Staffing level in CERN budget
External funding
LCG Phase 1
EGEE
Phase 1
EGEE Phase 2 (assumed)
Staffing required for planned service level
102 FTE-years missing for Phase 2

16. Progressive evolution
Preparing for 2007
2003 – has demonstrated event production
In 2004 we must show that we can also handle the data – even if the computing model is very simple
-- This is a key goal of the Data Challenges & Service Challenges
Target for end of this year –
Basic model demonstrated using current grid middleware
All Tier-1s and ~25% of Tier-2s operating a reliable service
Validate security model, understand storage model
Clear idea of the performance, scaling, and management issues
2004
2005
2006
2007
first data
Initial service in operation
Decisions on final core middleware
Demonstrate core data handling and batch analysis
Installation and commissioning
Progressive evolution

17. Running on several Grids
Aiming for common interfaces and interoperability

18. Conclusions
We are now in a position to start from an operational environment towards a computing system for LHC by
Completing functionality
Improving reliability
Ramping up capacity and performance
Data challenges and service challenges are essential to keep the right track
All partners are committed to arrive at interoperability of the grids involved
We are grateful for the very significant contributions from and the effective collaboration with the US
The full funding for LCG phase 2 still needs to be secured