GridPP Presentation to PPARC eScience Committee

1
GridPP Presentation to PPARC e-Science
Committee 26 July 2001
Steve Lloyd Tony Doyle John Gordon
2
Outline

• Component Model
• Resource Allocation and Funding Scenarios
• Intl Financial Comparisons
• Intl Grid Collaborations
• Grid Architecture(s)
• Links with Industry
• Summary
• Addendum
• 1. VISTA and GridPP
• 2. GridPP monitoring page

3
GridPP Proposal

• GridPP Vertically integrated programme
• component model...
• Input to development of 15-20M funding scenarios
  

4
GridPP Workgroups
Technical work broken down into several
workgroups - broad overlap with EU DataGrid
A - Workload Management Provision of software
that schedule application processing requests
amongst resources
F - Networking Network fabric provision through
to integration of network services into middleware
G - Prototype Grid Implementation of a UK Grid
prototype tying together new and existing
facilities
B - Information Services and Data
Management Provision of software tools to provide
flexible transparent and reliable access to the
data
H - Software Support Provide services to enable
the development, testing and deployment of
middleware and applications at institutes
C - Monitoring Services All aspects of monitoring
Grid services
I - Experimental Objectives Responsible for
ensuring development of GridPP is driven by needs
of UK PP experiments
D - Fabric Management and Mass Storage Integration
of heterogeneous resources into common Grid
framework
E - Security Security mechanisms from
Certification Authorities to low level components
J - Dissemination Ensure good dissemination of
developments arising from GridPP into other
communities and vice versa
5
Components 1-4 21M
6
20M Project
7.1m ? 6.7m
H 3.2
3.2 ? 2.9
7
17M Project
I 2.49m ? 1.2m
Experiment
7.1m ? 6.7m ? 6.0m
Objectives
H 3.2
3.2 ? 2.9 ? 2.45m
8
Experiment Objectives
50 reduction? 23 SY

• Vertically integrated programme?
• Broken component model
• Specific experiments or overall reduction?
• To be determined by Experiments Board

9
CERN (Component 3)
10 reduction? 3.1 SY
Basic Grid functionality UK-CERN integrated
programme - synergies, but cuts here will impact
10
CERN (Component 4)
50 reduction? 11 SY
Experiments support similar conclusions to
UK-based programme Non-UK funding dependencies?
15 reduction? 0.2M
HARDWARE
Pro-rata reduction on disk, tape, CPU...
11
Workload/Data Management
10 reduction? 1.2 SY
Reduced long-term programme? e.g. scheduler
optimisation (WG A) query optimisation
(WG B) or overall reduction?
12
15M Project
I 2.49m ? 0
Experiment
5m
Objectives
H 3.2
3.2 ? 2.9 ? 2.45m
13
15M Project

• Summary
• Even a 21-20M reduction is not trivial..
• EU DataGrid commitments are built in
• Focus on CERN and UK Capital as largest single
  items, then reduce workgroup allocations
• 17M budget cuts hard into the project
• Examples are based on original Component Model
• 15M budget is impossible within the Component
  Model
• A fixed allocation help in planning the start-up
  phase

14
International Comparisons

• PP Grids under development
• France
• Germany
• Italy
• US
• CMS
• Atlas

15
International Comparisons

• Summary - different countries, different models
• France Germany budget for hardware, assume
  staff
• Italy - lots of hardware and staff
• US - funds split between Tier1/2, Universities,
  infrastructure, and RD
• Italy gt UK France (EU) US (GriPhyN, PPDG and
  iVDGL characteristics within GridPP single UK
  programme)

16
GridPP Architecture
The DataGrid Architecture Version 2 German
Cancio, CERN Steve M. Fisher, RAL Tim Folkes,
RAL Francesco Giacomini, INFN Wolfgang Hoschek,
CERN Dave Kelsey, RAL Brian L. Tierney,
LBL/CERN July 2, 2001

• Based on EU DataGrid developments feeding into
  GGF
• Status Version 2 (2/7/01)
• Key elements
• Evolutionary capability
• Service via Protocols and Client APIs
• Representation using UML (TogetherSoft)
• Defines responsibilities of Work Packages
• Built from Infrastructure
• Based on PP Use Cases (applies to GridPP)

17
The Grid and Industry

• Help us develop the Grid
• Supply hardware - PCs, Disks, Mass Storage,
  Networking etc
• Supply software, middleware, management systems,
  databases etc
• Use the Grid for themselves
• Collaborative Engineering
• Massive simulation
• Federating their own worldwide databases
• Sell or develop the Grid for others
• Computation Services, Data services etc

18
Summary

• Balanced exploitation programme costs 21M
• 20M-17M-15M 3-year funding scenarios examined
• 20M maintains balanced programme
• 17M reduced experimental objectives
• 15M eliminates experimental objectives
• Final balance depends on funding allocation
• Emphasis on vertical integration component model
• International comparisons Italy gt UK France
  (EU) US (GriPhyN, PPDG and iVDGL characteristics
  within GridPP single UK programme)
• Contacts established with GriPhyN, PPDG and iVDGL
• InterGrid Co-ordination Group in development
• Architecture defined by GGF via lead in DataGrid
• Industry links emphasis on partnership

19
GridPP and VISTA

• Astrogrid will federate VISTA data with other
  large databases elsewhere
• this requires that VISTA data has already been
  processed and catalogues and images are
  available.
• VISTA have a proposal (e-VPAS) that concentrates
  on producing the databases on which the Astrogrid
  tools will work. This work has much in common
  with GridPP
• a similar timescale
• very large data flows from one remote site
• many distributed users
• reprocessing of data
• utilization of distributed computing resources
• GridPP have started discussions with VISTA and
  EPCC (GenGrid) as to how we can collaborate and
  share expertise and middleware

20
GridPP Monitoring Page

• Various sites now set up with UK Globus
  certificates
• Grid Monitoring
• Polls Grid test-bed sites via globus-job-run
  command
• Runs basic script producing XML encoded status
  information
• Load average and timestamp information retrieved
• Current status and archived load information is
  plotted...
• To be done...
• JAVA CoG kit being investigated (more robust)
• Simple monitoring system to verify test-bed
  timestamps (in case not everyone is using
  NTP)
• Integrate with the Grid Monitoring Architecture
• Incorporate current network bandwidth
  measurements into graphical system
• Automatic notification system