EGEE Organisation Steve Fisher / RAL 4-5/3/2004 This talk will cover a number of issues. The...

EGEEOrganisation

Steve Fisher / RAL4-5/3/2004

<[email protected]>

This talk will cover a number of issues.

The structure may be hard to discern

Steve Fisher/RAL - 4-5/3/2004EGEE Organisation 2

Credits• My slides have been stolen from many

sources including:– Fabrizio Gagliardi– Frederick Hemmer


Who’s who

G. ZaquineA. Aimar

I. BirdC. Vistoli

F. Hemmer?

M. AktinsonJ. Dyer

G. WormserF. Harris

F.BonnassieuxJ. Orellana

F. Hedman?

F. Gagliardi

B. Jones

A. BlateckyE. JessenT. PriolD. Snelling


Project Conferences• Twice a year for the duration of the project• Try to follow the EU presidential countries

– April ‘04 UCC Ireland (preceded by an Irish ministerial event on the EU research infrastructures in Dublin which EGEE will support within the NA5 activity)

– Nov ‘04 NIKHEF (3rd week proposed)– Mar/Apr Greece (skip the Luxemburg presidency

because we have no EGEE partners there)

– Nov ‘05 Scotland, no firm details yet


Future• EGEE seen by the EU as the most important

Grid infrastructure activity.

• Essential to be ready and "hit the ground running" on April 1st 2004 - Fab

• Second phase of the project to be proposed to the EU in 2005


JRA1: Objectives• Provide robust, supportable middleware components

– Select, re-engineer, integrate identified Grid Services– Evolve towards Services Oriented Architecture– Adopt emerging OGSI WS-RF standards– Multiple platforms

• Selection of Middleware based on requirements of– The applications (Bio & HEP)

• In particular requirements are expected from LCG’s ARDA & HepCALII

– The Operations• E.g. deployment, updates, packaging, etc..

• Support and evolve the middleware components– Evolve towards OGSI, or WS-RF or …– Define a re-engineering process– Address multiplatform, multiple implementations and interoperability

issues– Define defect handling processes and responsibilities


JRA1: Organization


LCG and R-GMA• LCG are now saying publicly that they want

R-GMA for:– Accounting– Network monitoring– Job monitoring (CMS)

• We have a draft plan being worked on with Ian and Zdenek– Dates in range Feb-Mar– Delays in completing the plan

• Where is source• How do RPMs get built etc…


EGEE• All middleware is one WP (JRA1, under

Frederic Hemmer)– Made up of 3 “coding” clusters

• Data Management• Info and Monitoring• Workload management

– Security decoupled again

• Old EDG application testbed will start with LCG-2– With R-GMA

• MDS free zone

– LCG will run with LCG-2 for all of 2004


JRA1: Milestones and Deliverables for the first year

MonthDeliverables

&Milestones

ItemLead Partner

M03 MJRA1.1 Tools for middleware engineering and integration deployed

CERN

M03 DJRA1.1 (Document) Architecture and Planning (Release 1) CERN

M03 MJRA1.2 Software cluster development and testing infrastructure available

CERN

M05 MJRA1.3 Integration and testing infrastructure in place including test plans (Release 1)

CERN

M05 DJRA1.2 (Document) Design of grid services (Release 1) CERN

M09 MJRA1.4 Software for the Release Candidate 1 CERN

M10 MJRA1.5 Integrated Release Candidate 1 enters testing and validation period (Release 1)

CERN

M12 DJRA1.3 (Software) Software and associated documentation (Release 1)

CERN


JRA1: Responsibilities

Partner(s) Activity Effort

CERN Data Management, testing and integration, overall coordination

32

INFN + Datamat

CESNET

Resource access, brokering and accounting

Logging and bookkeeping

22

CCLRC Information collection and retrieval 9

CNRS Test tools support group 2


What we have to do• We are supposed to have:

– Cluster manager– Architect– Sysadmin– QA and Doc – 1 Unit tester– 3 Developers

• We are also supposed to contribute to security (JRA3)• Three new people being recruited

– 2 Developers• Both job offers accepted

– 1 Integrator/sysadmin• Job offer accepted


Roles• Will not have 1:1 correspondence between

roles in proposal and people• All roles will need at least 2 people

– Probably need to move people around a bit– To provide coverage during illness and holiday– 2 heads are better than 1– To encourage team thinking


Cross JRA1 groups• Groups

– EMT • meeting every few weeks

– ATeam • Small ARDA team is making a start – but made up of the

old WP managers!

– QA• Nothing happened yet

– Security• Nothing happened yet

• No ITeam (yet)


How does ARDA project fit in

• Some points from the ARDA web site – dated 19th Feb


Main participants:• The basic middleware development team.

– Joint European-US team, managed as part of the EGEE project, but including participation of the VDT leadership.

• At least one team in each experiment– undertakes to take part in adapting/developing the analysis system of

their experiment to use the basic middleware...

• Providers of common HEP tools– such as POOL, SEAL, ROOT, GANGA and PROOF...

• Users within experiments – exploit the evolving distributed analysis systems and provide

immediate feedback to the developers.

• Service providers from regional centres– deploying early versions of middleware and analysis systems.

• Teams from other sciences is welcomed– such as those participating in EGEE,.


Structure• The ARDA Technical Forum, chaired by the

ARDA Project Leader, will bring together all of the participants.

• As far as possible the common decisions needed to enable the participants to work together (specifications, targets, priorities, timescales, ..) will be made by consensus in the forum.

• The project has the status of an LCG project “area”, with the project leader an ex officio member of the PEB.


People• The ARDA Project Leader is Massimo Lamanna. • Each experiment has appointed one person to

coordinate their activity in the project: – ATLAS: David Adams, – CMS: Lucia Silvestris, – LHCb: Ulrik Egede. – ALICE to be nominated shortly.

• The leader of the EGEE Middleware team is Frédéric Hemmer.

• The Project Leader has been asked now to launch the project, and prepare an initial plan of action.


“New ARDA document”• Taking longer than expected• Preparing APIs for each service to give to

GAG– Using C or Java (not yet decided) rather than

WSDL to explain things

• Because of concerns that this prototype spec may be hard to change, a lot of effort is going into getting it right– Therefore nothing to show yet– Will be ready before GGF10!


Tools and Procedures• CVS

– There will be a new CVS set up• probably just one• at CERN• ssh access• NOT on AFS

• Project “Portal”– EGEE will use Savannah (http://savannah.cern.ch/) for:

• Bug tracking• Communication• News• etc…

– It is very configurable with new modules being developed

• SCM draft from Alberto Di Meglio


JRA1 – SA1 interactions• LCG = SA1 ???• So far 1 meeting

– Zdenek Sekera, Frederic Hemmer, Erwin Laure, Peter Kunszt, Ian Bird, Maite Barroso, Marco Serra, Alberto di Meglio, Markus Shulz, Leanne Guy

• Some EDG issues:– Certification also included integration activities.

We want to clearly separate both– Clean border between JRA1 and SA1:

• tar file and makefile. • SA1 do not want packages in any format.

EGEE Organisation Steve Fisher / RAL 4-5/3/2004 This talk will cover a number of issues. The...

Documents

Transcript of EGEE Organisation Steve Fisher / RAL 4-5/3/2004 This talk will cover a number of issues. The...