INFSO-RI-508833
Enabling Grids for E-sciencE
www.eu-egee.org
EGEE is a project co-funded by the European Commission under contract INFSO-RI-508833
The EGEE project: building an international production grid infrastructure
Enabling Grids for E-sciencE
INFSO-RI-508833
• EGEE - what is it and why is it needed?• Middleware – current and future• Operations – providing a stable service • Networking – enabling collaboration• Summary
The material for this talk has been contributed by many colleagues in the EGEE & LCG projects.
It is heavily based on Bob Jones’ talk at UK AHM 2004.
Enabling Grids for E-sciencE
INFSO-RI-508833
Technological push?
• Is Grid technology merely a ‘funding opportunity’?• Is it an example of scientists wanting to do something
because they (just about) could?
• Is it in fact a technology driven activity, without any real purpose?
• Consider this diagram
Enabling Grids for E-sciencE
INFSO-RI-508833
Grids vs. Distributed Computing
• Existing distributed applications:–tend to be specialised systems–intended for a single purpose or user group
• Grids go further and take into account:–Different kinds of resources
Not always the same hardware, data and applications
–Different kinds of interactions User groups or applications want to interact with Grids in
different ways
–Dynamic nature Resources and users added/removed/changed frequently
Enabling Grids for E-sciencE
INFSO-RI-508833
What is Grid Computing?
• A Virtual Organisation is:– People from different institutions working to solve a common
goal– Sharing distributed processing and data resources
• Grid infrastructure enables virtual organisations
“Grid computing is coordinated resource sharing and problem solving in dynamic, multi-institutional virtual organizations” (I.Foster)
Enabling Grids for E-sciencE
INFSO-RI-508833
The terms of the problem
• Technological progress produces more sophisticated digital sensors (particle physics detectors, satellites, radio-telescopes, synchrotrons…)
• Much of science is therefore becoming increasingly “data-intensive”
• Huge amounts of data need to be analyzed by large and geographically distributed scientific communities
• Consequently, single computers, clusters or supercomputers are not powerful enough for the necessary calculations and the data processing
Result: access to large facilities is difficult and expensive for the scientific community, particularly in less favoured countries=> increase of the “electronic divide”
Enabling Grids for E-sciencE
INFSO-RI-508833
The Grid: a possible solution
• The World Wide Web provides seamless access to information stored in different geographical locations
• The Grid provides seamless access to computing power and data storage capacity distributed over the globe
• Relies on advanced software, called middleware:
– authenticates, authorizes and accounts (AAA)– understands and locates the data which the
scientist needs– distributes the computing processing to
wherever in the world there is available and useful capacity
– sends the results back
The name Grid was chosen by analogy
with the electric power grid
Enabling Grids for E-sciencE
INFSO-RI-508833
Challenges
• Must share data between thousands of scientists with multiple interests
• Must connect major computer centres, not just PCs (not P2P computing)
• Must ensure that all data is accessible anywhere, anytime• Must grow rapidly, yet remain reliable for more than a decade
• Must cope with different computer centres access policies
• Must ensure data security
Enabling Grids for E-sciencE
INFSO-RI-508833
Benefits
• Effective and seamless collaboration of dispersed communities, scientific first and then industrial
• Ability to run large-scale applications aggregating thousands of computers, for very wide range of applications
• Transparent access to distributed resources from your desktop
• The term “e-Science” has been coined to express these benefits
• In the vision of the “Knowledge Grid”, the Grid can act as unifying agent between applications and non homogeneous data
Enabling Grids for E-sciencE
INFSO-RI-508833
What are the characteristics of a Grid system?
Numerous Resources
Ownership by MutuallyDistrustful Organizations
& Individuals
Potentially FaultyResources
Different SecurityRequirements
& Policies Required
Resources areHeterogeneous
Enabling Grids for E-sciencE
INFSO-RI-508833
What are the characteristics of a Grid system?
Numerous Resources
Ownership by MutuallyDistrustful Organizations
& Individuals
Potentially FaultyResources
Different SecurityRequirements
& Policies Required
Resources areHeterogeneous
GeographicallySeparated
Different ResourceManagementPolicies
Connected byHeterogeneous, Multi-Level Networks
Enabling Grids for E-sciencE
INFSO-RI-508833
EGEE Overview• Goal:
– Create a world-wide production-quality Gid infrastructure for e-Science
on top of present and future EU Research Networking infrastructure
• Build on: – EU and EU member states major
investments in Grid Technology– International connections (US and AP)– Several pioneering prototype results– Large Grid development teams in EU
require major EU funding effort• Approach
– Leverage current and planned national and regional Grid initiatives and infrastructures
– Work closely with relevant industrial Grid developers, NRENs and US-AP projects
• http://www.eu-egee.org
Applications
Geant-NREN networks
Grid infrastructure
Enabling Grids for E-sciencE
INFSO-RI-508833
The Grid: networked data processing centres and ”middleware” software as the “glue” of resources.
Researchers perform their activities regardless geographical location, interact with colleagues, share and access data
Scientific instruments and experiments provide huge amount of data
The (Science) Grid Vision
Enabling Grids for E-sciencE
INFSO-RI-508833
In 2 years EGEE will:
• Establish production quality sustained Grid services
– 3000 users from at least 5 disciplines– over 8,000 CPU's, 50 sites– over 5 Petabytes (1015) storage
• Demonstrate a viable general process to bring other scientific communities on board
• Propose a second phase in mid 2005 to take over EGEE in early 2006
Pilot New
Enabling Grids for E-sciencE
INFSO-RI-508833
EGEE and LCG
EGEE builds on the work of LCG to establish a grid operations service
• LCG (LHC Computing Grid) - Building and operating the LHC Grid
• A collaboration between:– The physicists and computing specialists
from the LHC experiment – The projects in Europe and the US that
have been developing Grid middleware– The regional and national computing
centres that provide resources for LHC – The research networks
Enabling Grids for E-sciencE
INFSO-RI-508833
EGEE Activities
• 48 % service activities (Grid Operations, Support and Management, Network Resource Provision)
• 24 % middleware re-engineering (Quality Assurance, Security, Network Services Development)
• 28 % networking (Management, Dissemination and Outreach, User Training and Education, Application Identification and Support, Policy and International Cooperation)
32 Million Euros EU funding over 2 years started 1st April 2004
Emphasis in EGEE is on operating a productiongrid and supporting the end-users
Enabling Grids for E-sciencE
INFSO-RI-508833
• EGEE - what is it and why is it needed?• Middleware – current and future• Operations – providing a stable service • Networking – enabling collaboration• Summary
Enabling Grids for E-sciencE
INFSO-RI-508833
gLite
• “gLite” - the new EGEE middleware• Service oriented - components that are :
– Loosely coupled (by messages)– Accessible across network; modular and self-contained;
clean modes of failure– So can change implementation without changing interfaces– Can be developed in anticipation of new uses
• … and are based on standards. Opens EGEE to:– New middleware (plethora of tools now available)– Heterogeneous resources (storage, computation…)– Interact with other Grids (international, regional and national)
Enabling Grids for E-sciencE
INFSO-RI-508833
Architecture Guiding Principles
• Lightweight (existing) services – Easily and quickly deployable– Use existing services where possible as
basis for re-engineering• Interoperability
– Allow for multiple implementations• Resilience and Fault Tolerance
• Co-existence with deployed infrastructure– Reduce requirements on site components– Co-existence (and convergence) with LCG-2 and Grid3 are essential for the EGEE
Grid service
• Service oriented approach– Follow WSRF standardization– No mature WSRF implementations exist to date so start with plain WS (WS-I)– Provide framework to others so higher-level services can be developed quickly
Architecture: https://edms.cern.ch/document/476451
Enabling Grids for E-sciencE
INFSO-RI-508833
• EGEE - what is it and why is it needed?• Middleware – current and future• Operations – providing a stable service
– Needs more than middleware– Organisational, operational infrastructure
• Networking – enabling collaboration• Summary
Enabling Grids for E-sciencE
INFSO-RI-508833
User-view of EGEE: a multi-VO Grid
User Interface
Grid services
User Interface
Enabling Grids for E-sciencE
INFSO-RI-508833
EGEE: adding a VO
EGEE has a formal procedure for adding selected new user communities (Virtual Organisations):
• Negotiation with one of the Regional Operations Centres
• Seek balance between the resources contributed by a VO and those that they consume.
• Resource allocation will be made at the VO level. • Many resources need to be available to multiple VOs :
shared use of resources is fundamental to a Grid
Enabling Grids for E-sciencE
INFSO-RI-508833
SA1 - Operations
• Scale of the production service– April 2004: ~2000 CPUs over ~ 30 sites (LCG-1 → LCG-2)– December 2004: ~8000 CPUs over ~ 80 sites (Migrated to
Scientific Linux) This is far beyond the project milestones!
– Continuous improvements to LCG-2 middleware
• Set-up of CIC/ROCs– Roles/responsibilities defined in execution plans
documented and implemented
• On-going:– Complete set-up of pre-production service– Deployment planning for gLite (EGEE1 M/W version)– Deploy accounting infrastructure
Enabling Grids for E-sciencE
INFSO-RI-508833
Running the Production Service Grid deployment has entered a new phase• Basic middleware is working
– responsible now for a small fraction of the problems• Outstanding performance/functionality issues
– RLS, RB / little modularity & lack of consistent interfaces …– some solutions are being developed but many cannot be addressed in current
software/architecture - set priorities for new middleware (gLite)• Many operational issues
– mis-configuration, out of date mware, single points of failure, failover, mgmt interfaces …– resources unsuitable for applications needs (e.g. insufficient disk space)– slow response by sites to problems (holiday periods, security concerns)– new middleware will not help for many of these issues - grid partners must think Service
The grid still does not appear as a single coherent facilityapplications must adapt to the current service to gain maximum profit but result has been very effective for LHCb - ~3000 concurrent jobs
Enabling Grids for E-sciencE
INFSO-RI-508833
• Operation Management Centre– located at CERN, coordinates
operations and management– coordinates with other grid projects
• Core Infrastructure Centres– behave as single organisations– operate core services (VO specific
and general Grid services)– develop new management tools– provide support to the Regional
Operations Centres
EGEE Operations (I): OMC and CIC
Enabling Grids for E-sciencE
INFSO-RI-508833
EGEE Operations: ROC
• Regional Operations Centre responsibilities and roles:– Testing (certification) of new middleware on a variety of
platforms before deployment– Deployment of middleware releases + coordination +
distribution inside the region – integration of ‘Local’ VO– Development of procedures and capabilities to operate the
resources– First-line user support– Bring new resources into the infrastructure and support their
operation– Coordination of integration of national grid infrastructures
Provide resources for pre-production service
Enabling Grids for E-sciencE
INFSO-RI-508833
Production grid service
Launched Sept’03 with 12 sites, now more than 100 sites and continues to grow
Enabling Grids for E-sciencE
INFSO-RI-508833
Production grid service
Launched Sept’03 with 12 sites, now more than 100 sites and continues to grow
Enabling Grids for E-sciencE
INFSO-RI-508833
Production grid service
Launched Sept’03 with 12 sites, now more than 100 sites and continues to grow
Enabling Grids for E-sciencE
INFSO-RI-508833
Grid projects
Many Grid development efforts — all over the world
•NASA Information Power Grid•DOE Science Grid•NSF National Virtual Observatory•NSF GriPhyN•DOE Particle Physics Data Grid•NSF TeraGrid•DOE ASCI Grid•DOE Earth Systems Grid•DARPA CoABS Grid•NEESGrid•DOH BIRN•NSF iVDGL
•EuroGrid (Unicore)•DataTag (CERN,…)•DataGrid (CERN, ...)•Astrophysical Virtual Observatory•GRIP (Globus/Unicore)•GRIA (Industrial applications)•GridLab (Cactus Toolkit)•CrossGrid (Infrastructure Components)•EGSO (Solar Physics)
•UK e-Science Grid•Netherlands – VLAM, PolderGrid•Germany – UNICORE, Grid proposal•France – Grid funding approved•Italy – INFN Grid•Eire – Grid proposals•Switzerland - Network/Grid proposal•Hungary – DemoGrid, Grid proposal•Norway, Sweden - NorduGrid
Enabling Grids for E-sciencE
INFSO-RI-508833
Authentication, Authorisation
• Authentication
– User obtains certificate from CA– Connects to UI by ssh– Downloads certificate– Invokes Proxy server – Single logon – to UI - then Secure Socket
Layer with proxy identifies user to other nodes
• Authorisation - currently
– User joins Virtual Organisation– VO negotiates access to Grid nodes and
resources (CE, SE)– Authorisation tested by CE, SE:
gridmapfile maps user to local account
UI
CA
VO mgr
Personal
VO database
Gridmapfiles
On CE, SE nodes
SSL
(proxy)
VO service
Enabling Grids for E-sciencE
INFSO-RI-508833
JRA3 - EGEE Authentication Scheme- EUGridPMA
• Policy Management Authority: “Club” of trusted Certification Authority managers www.eugridpma.org
Green: CA Accredited Yellow: being
discussedOther Accredited CAs: DoEGrids (US) GridCanada ASCCG (Taiwan) CERN Russia (HEP) FNAL Service CA (US) Israel Pakistan
Greece: Hellasgrid CA (AUTH)
Enabling Grids for E-sciencE
INFSO-RI-508833
• EGEE - what is it and why is it needed?• Middleware – current and future• Operations – providing a stable service • Networking – enabling collaboration
– Current application communities
• Summary
Enabling Grids for E-sciencE
INFSO-RI-508833
Bringing new applications to the grid
1. Outreach events inform people about the grid / EGEE
2. Application experts discuss specific characteristics with the users
3. Migrate application to EGEE infrastructure with the help of EGEE experts
4. Initial deployment for testing purposes
5. Production usage - user community contributes computing resources for heavy production demands - “Canadian dinner party”
Enabling Grids for E-sciencE
INFSO-RI-508833
NA3 – User training and induction
• NA3 has been involved in more than 130 training events across the world– (including the GGF and other grid schools)– ~2000 people trained
induction; application developer; advanced; activity retreats– Material archive online with ~1000 presentations
• Strong links made with GILDA testbed and use of GENIUS portal– Regularly used as part of tutorials
• Essential element of the virtuous cycle for new communities– Training is one of the first things new communities need
• Process for handling feedback defined– Helping to improve material and organisation
• Roadmap for future event planned– Open to new suggestions– Produced status report and update training plan taking into account lessons
learned• On-going
– Plan for next EGEE M/W (gLite) training
Enabling Grids for E-sciencE
INFSO-RI-508833
EGEE User Support: infrastructure
General approach: 3 main support centers to guarantee coverage 24/7 and 365 day support and provide a single point of contact to customers and to local Grid operations.
To ensure 24x7 support, it was decided to have 3 GGUS teams in different time zones. GGUS started off at Forschungszentrum Karlsruhe in GermanyGermany in 2003 and has had a partner group at Academia Sinica in TaiwanTaiwan since April 2004. A third partner in North AmericaNorth America will complete the 24 hours cycle.
Enabling Grids for E-sciencE
INFSO-RI-508833
EGEE User Support: infrastructure The ROCs and VOs and the other project wide groups such as the Core
Infrastructure Center (CIC), middleware groups (JRA), network groups (NA), service groups (SA) will be connected via a central integration platform provided by GGUS.
This central helpdesk keeps track of all service requests and assigns them to the appropriate support groups. In this way, formal communication between all support groups is possible. To enable this, each group has to build only one interface between its internal support structure and the central GGUS application.
Enabling Grids for E-sciencE
INFSO-RI-508833
• More about applications and communities in the next talk
Enabling Grids for E-sciencE
INFSO-RI-508833
• EGEE - what is it and why is it needed?• Middleware – current and future• Operations – providing a stable service • Networking – enabling collaboration
– Current application communities– Enabling new and effective use of EGEE
• Summary
Enabling Grids for E-sciencE
INFSO-RI-508833
Who else can benefit from EGEE?
• EGEE Generic Applications Advisory Panel:
– For new applications
• EU projects: MammoGrid, Diligent, SEE-GRID …
• Expression of interest: Planck/Gaia (astroparticle), SimDat (drug discovery)
Enabling Grids for E-sciencE
INFSO-RI-508833
Intellectual Property• The existing EGEE grid middleware (LCG-2) is
distributed under an Open Source License developed by EU DataGrid
– Derived from modified BSD - no restriction on usage (academic or commercial) beyond acknowledgement
– Same approach for new middleware (gLite)
• Application software maintains its own licensing scheme
– Sites must obtain appropriate licenses before installation
Enabling Grids for E-sciencE
INFSO-RI-508833
Summary
• EGEE is the first attempt to build a worldwide Grid infrastructure for data intensive applications from many scientific domains
• A large-scale production grid service is already deployed and being used for HEP and BioMed applications with new applications being ported
• Resources & user groups will rapidly expand during the project
• A process is in place for migrating new applications to the EGEE infrastructure
• A training programme has started with events already held
• Prototype “next generation” middleware is being tested (gLite)
• Plans for a follow-on project are being discussed
Enabling Grids for E-sciencE
INFSO-RI-508833
When will the grid disappear?
• Two possibilities:
• 1. Grids will not fulfill their promise and fade into being a niche distributed computing domain
• 2 Grids will become ubiquitous and easily usable – transparent to the user and so ‘disappear’– Following the trajectory of other networked services