Meeting of 3rd March 2009, Catania. Bob Jones, Steven Newhouse EGEE Ludek Matyska, Jürgen Knobloch, Dieter Kranzlmüller EGI Hermann Lederer, Wolfgang Gentzsch DEISA Leif Laaksonen, Martin Polak, Peter Kunszt PRACE Hans Döbling, DANTE David Foster, TERENA Meeting chair: Bob Jones Notes: Peter Kunszt The meeting started with each participants introducing themselves, their affiliation and what value they see in holding this meeting. We then went into a discussion about the purpose of this group: Do we need to meet regularly? - Mandate? HL: European projects here with specific aims, that are complementary and/or overlapping What are current problems in each project that is relevant to others, how do we go about it, what's next? Role of networks is not clear yet, DEISA has a strong link already. Networks will become more important JK: Many people have several 'hats' - very common in EGI. Overlap of institutions, people on the projects.. Find convergence. Middleware, accounting, etc LM: understanding needed on roles BJ: What is our vision? How do they match - work on management level Community support already works together with DEISA (fusion) User communities have to drive this. LL: Mandate - recommendations by this group should also have influence on EC policies HD: We are here because someone else made a 'master plan' (layered e-Infrastructures) Understand interfaces - technology, identity (pma-policies) DK: every organization needs to fulfill its own obligations toward its users, and has to find the funding for it LL: we can all influence it. LM: maybe deliverables can be shared before the result is published - Composition? PK: The members of each research infrastructure are here: EGI, PRACE (in construction with the help of EGEE and DEISA), TERENA, DANTE How do they interact, who does what - how to exploit each others competence, synergies Documents available: EGI blueprint, soon also PRACE ecosystem analysis (after review this week) WG: Collaboration is needed between these infrastructures for the sake of our end-users - Other parties needed? No obvious candidates were identified - Scope? * European, do we want to keep it like that? * Having another meeting with US and Asian colleagues? Sharing Information. - Preparation of strategies, strategic documents should be done such that they are commented by relevant other parties - Structures, long term commitments of each project should be communicated to each other x sustainability x technical feasibility x usage models Form common opinions Further develop vision for future ecosystem Communicate with other similar structures HD: keep the scope in europe WG: it is too ambitious to talk to everyone HD: and we don't have a mandate LM: we would not even know whom to ask for one At this point Peter Kunszt had to leave the meeting to go to the airport. Based on the above discussion we identified the following objectives of this group: - Share information within the group on status/direction of our infrastructures - Gather input and form common opinion on subjects that will impact the long-term goals, function and effectiveness of the e-Infrastructure ecosystem - Present an agreed European position on e-Infrastructure to peer infrastructures elsewhere in the world - Further develop the vision and model for Europe's e-Infrastructure - Be able to offer a clear added-value to key user communities (e.g. future ESFRI RIs) where we know they will need networks/grids/supercomputers and help them devise computing models that would make use of the European e-Infrastructures - Interact with future structures (e.g. future internet, commercial cloud systems, green IT) to highlight the lessons learnt from existing production infrastructures and formulate common user requirements We also discussed how the networking models for PRACE and EGEE may differ and the relevance of the LHC network model for other ESFRI RIs (noting that some are more like sensor networks). At the firstmeeting we had considered PRACE sites as data sources and we should further build on this idea to define the ecosystem vision which can be documented and given to the ESFRI RIs as a potential model. This group should meet 3 to 4 times per year and each infrastruture/project will take turns in hosting the meeting, preferrably piggy-backing on some existing event to reduce additional travel. Telephone conference facilities should also be available so members that can join remotely. Membership of this group should be limited to the management representatives so that we can discuss openly and freely. It can cause joint-technical groups to work together to address specific issues which will advance the mandate/vision when driven by well identified user requirements. We also agreed that we will make more use of common dissemination channels for advertising the work of e-Infrastructures. An example given is the iSGTW weekly online magazine (http://www.isgtw.org/) which started with EGEE but is now supported by the GridTalk project. iSGWT has already carried articles about GEANT and DEISA applications and is willing to do more of this in the future. Similarly, we will co-locate an event for our infrastructures (similar to Catania where the Open Grid Forum meeting and the EGEE user-forum are co-located) in the 2010 timeframe. Based on these objectives, we prepared a set of actions to drive the agenda for the next meeting, to be hosted by DEISA/PRACE in Amsterdam on 11, 12, or 13 May. Actions: 1. Set-up doodle to schedule the exact date/time of the meeting in Amsterdam (Hermann) 2. Convert the above objectives into a mandate/vision statement to be finalised at the meeting and disseminated widely (Bob) 3. Each infrastructure will provide material as input to the next meeting about the contacts it has made with the ESFRI projects and the methods for interaction it has used. (EGEE/EGI/DEISA/PRACE/GEANT) Agenda for meeting in Amsterdam: - Finalise mandate and vision document to be disseminated widely - Pool input (provided before the meeting) from each infrastructure to build as complete a picture as possible about the e-Infrastructure needs of the ESFRI projects and the contacts already established. Analyse this information to establish a priority list for which ESFRI projects we should interact with. Consider if the tentative GEANT3 event in 2009 could be an opportunity to interact with some of the high-priority ESFRI projects. - schedule further 2009 meetings