PEB meeting

600/R-002 (CERN)



Show room on map
Here is the proposed agenda for the PEB EGEE PEB meeting on Friday 22nd August room 600-R-002 at CERN. It is split into two parts: the first is for the PEB members (this will be the first occasion they meet) and then the second part is for the work on the technical annex and includes other people contributing to the document. A telephone conference has been organized: +41 22 767 7000 ask for EGEE PEB (organizer: R.Jones)
    • 9:00 AM 10:00 AM
      PEB only

      09:00-10:00 PEB members only
      Brief introduction of the mandate of the PEB: Fab
      Discussion on the mode of working of the PEB for the preparation of the TA draft and through to the end of the negotiation process
      Brief status of the TA draft and timeline for meeting the 1st September deadline: Bob

      EGEE PEB meeting 22nd August 2003 Present: Fabrizio Gagliardi (Project director) Bob Jones (Technical director) Guy Wormser (Applications) Frederic Hemmer (mware re-engineering) Fredrik Hedman (security) Ian Bird (grid operations) Excused: Gabriel Zaquine (quality assurance), Malcolm Atkinson (dissemination) This was the first meeting of the EGEE PEB and Fab reminded the participants of the mandate of the PEB and how it operates within the overall project management structure. PEB mode of operation: PEB should meet weekly until end of October to ensure successful negotiation process then move to a meeting every two weeks until the end of the 2003 and reassess the situation for January-April. Once the project has started (April 2004) there will be weekly meetings. A telephone conference will be organized for each meeting and minutes will be kept along with an action list. If PEB members cannot participate in a PEB meeting they should make this know to the Project Office beforehand and send a deputy sufficiently familiar with the status of the activities concerned to provide useful input and be empowered to take decisions. Meetings will be a maximum 2 hours long - the regular slot will be 15:00 to 17:00 on Thursday but this may be modified exceptionally. At CERN will be held in 600-r-002. The PEB are expected to interact daily during the preparation of the TA. Status of TA: Bob summarised the situation of the TA During the discussion the following points were clarified: Russia will be considered as a federation and we need to clarify how the cut of 1Meuros will be distributed across NA and SA activities. Discussed how to handle the 4.5Meuro budget cut for SA1. Ian and Bob will distribute an updated excel table as requested by the executive committee. At Proposal time we used the concept of virtual FTEs (100Keuros/year per FTE) but in the TA we need to be more specific in terms of true budget including unfunded contributions. Priorities for TA: answering questions raised by EU, essentially SA1, JRA1 and dissemination. Next PEB meeting: Wednesday 27th August 17:00-18:00 to review the status of the TA.
    • 10:30 AM 12:00 PM
      Technical Annex

      10:30-12:00 PEB + Technical Annex contributors
      Issues arising with TA draft:
      Networking activities:
      NA1 & NA5: Fab
      NA2: John Dyer
      NA3: Mark Parsons/Malcolm Atkinson
      NA4: Guy Wormser/Vincent Breton/
      Grid operations:
      SA1: Ian Bird
      Joint Research Activities:
      JRA1: Frederic Hemmer
      JRA3: Fredrik Hedman
      Network related activities: SA1/JRA4: Paul Bright-Thomas/Franck Bonnassieux/Peter Clarke

      EGEE Technical annex meeting. Friday 22nd August 10:30-12:00 Present: John Dyer (phone) Paul Bright-Thomas (phone) Mark Parsons(phone) Fab Gagliardi Bob Jones Guy Wormser Frank Harris Vincent Breton Ian Bird Franck Bonnassieux Review of the status of material for the Technical Annex (TA): NA1 (Fab): first draft sent identifying key people of PEB. Still to do figure updates to show split of dissemination/application PEB role into two. NA5 (Fab): More difficult. A catch-all for networking activities. Proposal text was joint effort of Fab & Mark. Resources are scarce in NA5 - the 3 FTEs listed for CERN are really intended for NRENs. A phone conference will be held this evening on this subject with NRENs, DANTE & GEANT. Once this is clarified it will need to be coordinated with other NA activities. NA2 (John): Revision sent to Mark Parsons addressing most of the appropriate EU points. Still need to address concertation issues - not clear if coordination or management is necessary. Decision: concertation are better placed in NA5 rather than NA2. Still need to clarify the management structure for NA2/3/4 between NeSC& Terena so text will be updated as soon as there is a common understanding. Action: John, Guy, Mark to have a phone conference today NA4 (Guy): 3 main applications: HEP (Frank Harris), Bio (Vincent Breton), generic (Roberto Barbera). Guy will organise a phone conference starting Monday in particular the EU question on policy for resource allocation. Text will expand on procedure for introducing new VOs and how the allocation policy will work. A more detailed description of Bio activities has been prepared by Vincent. Input from Russian institutes also included. A complication is to replace the reliance on HEAVEN (failed project proposal) with some EGEE coordination work. Frank and JJ Blaising have investigated the issues for HEP concerning NA4 deliverables and how they fit in with the LHC experiments milestones and plans. Frank is consulting with Federico Carminati and they are using one of the LCG groups, GAG, to get feedback from the experiments. Frank will also interact with the experiments' computing coordinators and will try to make it compatible with the HEP input on mware and grid operations. Fab suggested the LHC experiments' architects should also be involved. In terms of FTEs, 4 are allocated for HEP at CERN. Fab clarified that CERN (not other institutes in the experiments) must match these 4 FTEs to create a team of 8. The TA text for 1st September does not need to identify the individuals or the exact structure that will be used for organising this team but it will be necessary for subsequent drafts. Christian Sagay (CS) will also be consulted for industrial aspects. First draft for NA4 text will be available end of Monday. SA1: A first draft is ready (essentially place-holders). An important issue is the cut in budget (4.5Meuros) but Ian is waiting for input from federations. Another issue is how VOs join. Deliverables issues are not so difficult. Draft is likely to be too long in length compared to the guidelines - but this is natural for an activity that represents approximately 50% of the budget. Fundamental issue - "grid of grids" has a management basis but makes no sense from an operations point of view. In the discuss it was clarified that EGEE will re-engineer and deploy a single set of mware. An objective is to have inter-operating international grids but is more likely to be at the EGEE level. Need to clarify with Frederic Hemmer and Gabriel Zaquine (QA) the sw release process. Expect 2nd draft early next week. JRA1 (Frederic): current text is proposal text with removal of Russian institutes (as specified by EU). Issues are being discussed with people as they return from holidays. Security (JRA3) issues have been clarified with Fredrik Hedman. Frederic has taken input from the HEP GAG group (he went to the GAG meeting yesterday). HEP GAG would favour having separate implementations of grid services. While it is understood that EGEE will not be only source of mware, EGEE has to offer a defined level of service and hence will only deploy elements of satisfactory quality that do not cause disruption to services and have an adequate support structure. The starting point for EGEE mware will be the LCG sw suite. Expect a draft at the beginning of next week. JRA3 (Fredrik): roles between JRA1 and JRA3 have been reasonably clarified. Security will be similar to quality - people as foreseen in JRA1 to ensure the mware re-engineering does address security issues and comply to agreed security architecture. A cluster manager will be designated for this part of the work. All partners have been contacted except Norwegians (Bergen). How will application security requirements be taken into account? Can start by taking into account DataGrid deliverables and input from LCG then need a process for allowing feedback on security issues resulting in updated requirements lists. Expect to have a JRA3 draft for Wednesday. JRA2 (Guy for Gabriel): The intention is to have a good overlap with LCG for what concerns tools used as part of the software management and quality control. Alberto Aimar is responsible for this work in LCG and will discuss with Gabriel next week (Gabriel will be at CERN the whole week). One proposal is to give one of the French FTEs foreseen for JRA2 to JRA4/SA2 activities but this needs to be clarified next week. Alberto has prepared some input on QA matters starting from the proposal text. Hope to have a draft for Thursday. SA2/JRA4 (Franck): PBT met with DANTE for the issue of cooperation between the projects on these activities. PBT has seen a draft of the GEANT2 bid document that does include FTEs foreseen for activities closely related to EGEE activities. DANTE is likely to have 1 FTE for EGEE activity work. EGEE settled on 4 FTEs for these activities. Should write in TA that the relation between NRENs, GEANT and EGEE will be regulated by a MoU being prepared by the projects and that the MoU will be made available to the EU when ready. IPv6 deployment is foreseen in DANTE & GEANT2 but not clear how we can expand on this in TA since we cannot guarantee its availability. The approach to take is to include in JRAs text on how IPv6 features could be exploited for performance, guaranteed bandwidth, security etc. but without mentioning Ipv6 requirements in the deliverables. Failure in the deployment of IPv6 could also be mentioned as a potential risk for the JRA activities. Franck will distribute IPv6 features list to TA people so they can write text taking in account these features. For the last mile connectivity costs, in SA1 say a pre-requirement for connecting to the grid is that the centres must have sufficient network connections. UREC will lead SA2. UCL will lead JRA4. Unfunded partners are non-signatories (i.e. did not submit A2 form). The approach is for such organisations to contact the appropriate activity manager. If agreement is found then they can be reported as unfunded extra effort in quarterly report. Note: Such non-signatory manpower is additional and does not replace the unfunded contributions to be provided by the project partners. Such non-signatories that are already known can be added to table 1b of the TA. The 3 FTEs in JRA4/SA2will be moved from CERN for this work. Paul is drafting JRA4, Franck is drafting SA2 and they will be made coherent and available for Wednesday. Bob added that any new text should be submitted via the appropriate PEB member. In the case of JRA4 this will be Frederic Hemmer and for SA2 it will be Ian Bird. If PEB members have anyone they would like to add to the TA email list (used to notify people of changes) then they should send their email address to Bob.