An e-Infrastructure in Europe: a strategy and policy driven approach - PowerPoint PPT Presentation

1 / 28
About This Presentation
Title:

An e-Infrastructure in Europe: a strategy and policy driven approach

Description:

Significant amount of DIY will be increasingly required ... GRID Projects oriented services to users (EGEE, DataGrid, DEISA, SEE-GRID, many others) ... – PowerPoint PPT presentation

Number of Views:28
Avg rating:3.0/5.0
Slides: 29
Provided by: danyvan
Category:

less

Transcript and Presenter's Notes

Title: An e-Infrastructure in Europe: a strategy and policy driven approach


1
An e-Infrastructure in Europea strategy and
policy driven approach
  • Dany.Vandromme_at_renater.fr

2
Remind from Athens meeting
  • Progress with telecommunication industry rather
    than against! But
  • Significant amount of DIY will be increasingly
    required
  • NREN organisation is getting stronger and better
    organised
  • Sustained political support must be maintained

3
From Athens to Roma!
  • e-Infrastructure is not just physical networks.
  • e-Infrastructure is also services, like GRIDs,
    communities, access to other research
    infrastructures or just a top-quality Internet to
    access information.
  • Further global development is not only technology
    driven Need to set up the basis of a policy and
    strategy-driven approach to answer the user needs

4
EUROPEAN CONTEXT - NREN
  • European NREN Consortium made of 30 Members
    representative of countries NRENPC
  • Collective action for the deployment of the
    successive pan-European backbone network
    infrastructures
  • Policy driven decision making rules put in place
    by the NREN Policy Committee
  • Coordination and operational tasks delegated to
    DANTE, a not-for-profit company owned by the
    NRENs
  • Infrastructure cofunded by EC and NRENs

5
EUROPEAN CONTEXT - NREN
  • Back-up
  • NRENs are installed by the Member States ( other
    countries), which support their activity.
  • NRENs policies are defined and implemented
    nationally
  • NRENs collective action (like GEANT) is
    supported by the EC (via co-funding), and matches
    with the EU policy and startegic objectives.

6
EUROPEAN CONTEXT - NREN
  • Current achievement is GEANT network, which
    brought back Europe on the forefront of the world
    scene of Research Networks
  • More than the technology, the organisational
    model of GEANT is the more significant
    achievement
  • It allows Europe to expand its influence in other
    world regions (Med, LAC, AP), but also to revise
    the cost sharing habits for intercontinental
    connectivity!
  • Dedicated to serve all scientific communities
    (including but not only GRIDs) with shared AUP
    rules

7
EUROPEAN CONTEXT - GRIDS
  • GRID Projects oriented services to users (EGEE,
    DataGrid, DEISA, SEE-GRID, many others)
  • Development projects oriented towards
  • Scientific communities with wide coverage of
    services like grid computing, distributed
    storage, on-demand resource allocation (EGEE)
  • Usages for targeted users (like High Performance
    Computing with DEISA)

8
EUROPEAN CONTEXT - GRIDS
  • Virtual organisations (VO) for Research
  • Astronomy (VLBI, Virtual Observatory, Stellar
    databases, etc)
  • Climate (European Computing Facility, Real time
    data assimilation)
  • Earth observation (Data acquisition and merging)
  • Biology (today distributed collaborative
    databases, to-morrow on-line distributed instant
    computing)
  • Cooperative engineering (Aero, Auto, etc)

9
EUROPEAN CONTEXT - GRIDS
  • Other virtual organisations (VO) which may be
    found outside the RE world
  • Health care
  • Public services and administrations
  • Cooperative arts and cultural heritage
  • etc

10
EUROPEAN CONTEXT - GRIDS
  • Other projects oriented IT technology development
    like 6NET many national initiatives (IT, GR,
    DE, HU, UK, etc) numerous FP6 projects still
    in preparation)

11
EUROPEAN CONTEXT - GRIDS
  • Back-up
  • GRID projects are issued from FP (or national)
    project calls
  • When selected, they are supported by EC and the
    project partners
  • All projects are dedicated to develop and/or to
    serve (communities) but not from a national
    policy decision (even if they are welcomed)

12
ESFRI (1/2)
  • European Strategy Forum for Research
    Infrastructures
  • Created as a follow-on action from the RI
    Conference in Strasbourg in September 2001.
  • Made of high level government representatives of
    the EU Member States EC
  • Address all science fields in the context of
    Research Infrastructures of European significance

13
ESFRI (2/2)
  • Marine Research, Neutron sources, Free Electron
    Lasers, etc
  • Working Group on High Performance Computing and
    Networking since June 2003
  • Inventory of available resources across Europe
  • Identify science cases of which progress depend
    on HPCN development
  • Propose strategic guidance to progress

14
Other bodies and fora
  • ESF, EURAB, EIFORUM may happen to discuss about
    networks, grids and related topics.
  • International cooperation (NA, AP, LAC, AU, etc)
    both for infrastructure, middleware,
    normalisation and standards at all levels of the
    eInfrastructure.
  • Interactions with the non-RE world!

15
How does this work? (1/4)
  • 1) Variations about access?
  • Compliance with an AUP (GEANT)
  • Suited to the project objectives (DEISA)
  • Membership of the project itself (EGEE, 6NET) or
    of a predefined community (LCG)
  • ?

16
How does this work? (2/4)
  • 2) Scientific Excellence?
  • Scientific criteria for admission the
    scientific bodies, within projects/organisation
    are competent for evaluating the
    project/organisation itself, but not the
    scientific activities of users (except for
    community grids like LCG)
  • or objective scientific evaluation is made by the
    funding authorities.

17
How does this work? (3/4)
  • 3) Variable geometry and geographical coverage?
  • GRID infrastructures (including research
    networks) are a unique opportunity to integrate
    all countries in a European Research and
    Education Area
  • Networks are also THE way for all researchers to
    access large scale shared facilities in Europe
    and worldwide.

18
How does this work? (4/4)
  • 4) Information dissemination?
  • NRENs are not known from most users! Internet is!
    And users require very high quality Internet
  • GRID concept is unknown from almost everyone in
    Europe. Most researchers ignore what is (or could
    be) made available to them. But only the service
    needs to be properly advertised to users!
  • Surprisingly, big companies are marketing
    on-demand computing to the commercial without
    using the word GRID!

19
Requirements for a policy approach (1/4)
  • Cooperation between networks and services layers,
    organisation-wise
  • Cooperation at the technical and operational
    level a GRID service may require non-standard
    network resources in a completely automated manner

20
Requirements for a policy approach (2/4)
  • Network infrastructures are expected to be
    informative, but reactive and adaptive as well
  • GRIDs infrastructure may not expect using non
    existing network resources
  • There are a huge potential for sharing resources
    between GRIDs and networks (NOC, monitoring
    tools) to avoid useless duplication

21
Requirements for a policy approach (3/4)
  • Need to implement a user oriented approach for
    the presentation of services Most users do not
    need GRID services, while they need network
    services.
  • European integration will not occur without the
    full implementation of end-to-end services
  • eInfrastructure must be also a vector to
    prevent/fight the digital divide

22
Requirements for a policy approach (4/4)
  • Require to articulate the network and GRID
    initiatives/projects with the national and
    European science programmes and policies
  • In particular, there is no global resource
    allocation policy, able to combine peer review,
    funding constraints, national policies etc
  • Global discussion, political support etc

23
Rationale for a policy approach
  • E-Infrastructure is neither just networks, nor
    GRIDs, nor services It is all!
  • Avoid duplication between projects
  • Share information about available resources, best
    practices, service architecture, access, etc.
  • Standardisation of service interfaces for users
  • Based on long term strategy or vision (longer
    that the project timescales)

24
END OF PRESENTATION!
25
Future trends
  • Not to substitute to telecommunications
    operators, but acquire products and services not
    off-the-shelves.
  • Cooperate with Telcos but convince them to
    provide raw capacity at the lowest rates as
    possible and bring most of the technology
    mastering in the NRENs hands WDM instead of SDH,
    lit fibres instead of WDM, and eventually dark
    fibres instead of lit fibres when feasable.
  • Customize and provide network services by the
    NRENs themselves (DANTE for collective actions)

26
Challenges
  • Not to be destructive for the telecommunications
    industry, but incentive for new services
  • Coordinate disparate actions like xx-Light
    initiatives which promote lambdas rather than
    usages
  • Account for differentiated economical and
    regulatory contexts to harmonize the European
    network and reduce the digital divide.
  • TEN-155/GEANT issue was to managed simultaneously
    liberalized markets and monopolies Next will be
    to manage also markets with no rules!

27
Facts
  • It is acknowledged that TEN-155 and GEANT have
    significantly improved the research network
    situation in Europe.
  • However, the gap between extremes has also
    significantly increased.
  • Today, SEE countries are among the most expensive
    countries from GEANT point of view!
  • Poland and Czeck Republik have adopted a
    different approach to be among the most advanced
    partners.

28
Facts
  • The difficulty, reported for the SEE is well
    confirmed by the EUMEDCONNECT tendering process!
  • Better to cooperate (eventually being VERY
    incentive) with telcos than to fight them to
    improve the market situation, induced by the lack
    of competition.
  • Governments may help, because of specific needs
    for Education and Research.
  • New European rules for procurement will help.
Write a Comment
User Comments (0)
About PowerShow.com