GridPP PMB Minutes 6 - 15th October 2001 ========================================= Present: Roger Barlow (RB), Dave Britton (DB), Nick Brook(NB), Pete Clarke(PC) Tony Doyle(TD), Neil Geddes(NIG), John Gordon(JCG), Steve Lloyd(SLL) Robin Middleton(RPM), Les Robertson(LR) ************************************************************************* ***** IMPORTANT NOTE: Future PMB video conference will use *399876 * ************************************************************************* Review of Actions ----------------- ONGOING... 1.11 PC & DB to determine which project management tools to be used Will be done but not critical path. Not before 19th. CONTINUING 2.5 RPM to put place holder for InterGrid with potential interest from BaBar/CDF/D0 using prototype Tier-1/A at RAL. DONE. RPM presented BaBar case on Sunday. D0 status was presented by Kors Bos. CDF are discussing Grid development with D0/SAM. NIG and TD also present at InterGrid. TD presented GridPP 5' summary at that meeting. CLOSED 2.6 RB to track options via InterGrid from application point of view CLOSED: RB AND RPM TO CONTINUE TO LIASE ON INTERGRID (or High Energy and Nuclear Physics InterGrid Coordination Board - HENPIG) 3.2 PC, NB, RB + EB to define experimental requirements CONTINUING: SEE TIMELINE BELOW 3.8 NIG to provide details of Oversight Board timetable CONTINUING: MEETING WILL BE IN FIRST 2 WEEKS OF DECEMBER, PREFERABLY AFTER DECEMBER 7TH PEER REVIEW. MEMBERS WILL BE PAUL JEFFRIES, JOHN DARLINGTON, AND DARREN GREEN. PMB REQUESTS TERMS OF REFERENCE AND DATE. 3.10 DB to assess current financial commitments prior to request for transfer of up to 200k (total) via RAL SLA for travel this year. COMPLETE. APPEARS THERE IS SOME FREEDOM/DESIRE TO TRANSFER SOME OF HARDWARE COSTS TO NEXT FY SO PLENTY OF FLEXIBILITY EXISTS. 4.1 TD to set up the PMB web pages CONTINUING 4.2 NB to set up the EB web pages COMPLETE 4.5 RB to table a 1-page docuument to InterGrid on BaBar interest COMPLETE 4.7 RPM to formulate guidance on costs of sending person to CERN on LTA. Some examples would be useful. Agree with Ken Peach and circulate to PMB (Les would use this information to inform potential applicants). IN PROGRESS: COSTS INDEPENDENT OF GRADE; THREE SCENARIOS REQUESTED 4.9 JCG to prepare the advert in the EU Journal. COMPLETE 5.1 JCG and LR to discuss CERN hardware purchase. COMPLETE 5.4 PC to appoint deputy COMPLETE: ROBIN MIDDLETON APPOINTED. In addition to Robin's tasks as Deputy Technical Board Chair, DataGrid-UK Project Leader, and GridPP Budget Holder, he is now chair of the DataGrid Project Management Board (appointed for 1 year) [see below]. 5.5 RPM to report on current status of EU funded posts COMPLETE: INFORMATION CIRCULATED BY EMAIL. 5.6 JCG to estimate manpower needs for Tier 1/A centre CONTINUING: FLAGGED AS CRUCIAL INPUT NOW REQUIRED FOR TOP-DOWN PLANNING. 5.7 PC to come up with a list of names for TB CONTINUING: DIDN'T GET THIS FAR; MAY BE COMPLETE. 5.8 NIG to pass costings of the current allocated post to DB COMPLETE Project Technical Definition ---------------------------- PC has received input from experiments via EB (and commends their effort and flexibility). Technical board members currently reading the input and will prepare written input to PC and plan to meet on 17th October. PC will prepare first draft of Technical Assessment Document by 19th October. TB will meet on 23rd October before CB and a final version of the document is planned for about 26th Oct. PMB will then identify a list of approved project tasks and produce an invitation to Tender by 2nd Nov. The tender call will contain pointers to EB (one document per experiment please!) and TB document. Tender invitation will be extended to Institutes and a response invited by 30 Nov. Recruitment Timetable (for reference) --------------------- 8th October - EB input deadline. TECHNICAL INPUT ~COMPLETED 8TH OCTOBER. 19th October - Technical assessment document - first review of CERN posts 26th October - PMB definition of tasks 2nd November - Tender document 5-6th November - Collaboration Meeting 30th November- Deadline for submission 7th December - PRSC mtg. 14th December- summary document submitted to PPARC outcomes communicated to groups - second review of CERN posts COGS and GLOBUS ---------------- TD announced that the COGS initiative has been withdrawn; Globus will be put forward as the umbrella organisation for middleware. RPM expressed concern about loss of identity for DataGrid and GridPP. TD stated that he did not see GridPP having it's own CVS repository and all GridPP contributions would enter via DataGrid or Globus or experimental-specific repositories. General feeling that it was a good thing for GridPP to be seen to contribute to Globus and, in particular, no work should be delayed whilst this issue was debated (reference to Alex Martin's work). CERN HARDWARE ------------- There may be problems tendering for CERN hardware via the UK as some smaller suppliers may be put off by the necessity of offering site support at CERN. It might also be advantagous to simply allow CERN to do it and transfer the money to CERN later (possibility of slipping it out of this PPARC FY to help with the cash-flow problem). UK HARDWARE ----------- Tender notification has been issued and tender documentation under preparation. Hardware specifications to be circulated to EB. TOP-DOWN PROJECT PLANNING ------------------------- An initial discussion of top down project planning: TD proposed that we now adapt the submitted 17m project plan to increase the support to the experiments and try to target middleware support through DataGrid and Displacement. This was generally accepted. TD suggested a starting point of 6 posts for future and 6 posts for running experiments, with the balance of 3 posts left open pending input on Tier1/A support. Further discussion ensued. The approach needs to be agreed by the PMB by 26th October. DISPLACEMENT ------------ The RAL input from Dave Morrell is being incorporated into Dave Britton's financial model. It seems possible that ~2/3 of displacement could occur here. The balance would then be available to respond to downstream problems e.g. after the next grants round. MISCELLANEOUS ------------- Robin Middleton is now Chair of the DataGrid Project Management Board for the next 12 months. PROGRESS REPORT ON DATAGRID TESTBED (John Gordon, added following meeting) ----------------------------------- The DataGrid testbed will be publically announced at a meeting at CERN on 29th October. To ensure UK involvement and gauge the takeup John polled the potential Tier2 centres to check who will have resources ready to go. The replies were Liverpool - 6 Compass disk servers ready to act as Grid front-end to MAP but MAP interface still needs work. Not obvious they will be ready by mid November. Girish Patel will be at CERN on 29th. Imperial - 90 cpu cluster and 2.7TB of disk arrived last week. Should be ready. Dave Colling will attend. Edinburgh/Glasgow - ScotGrid Cluster+Datastore not delivered yet. Small Beowulf cluster available for now but network addressing problems. A 'maybe'. Cambridge - Andy Parker will have use of a Linux cluster and wants to be involved (perhaps not formally as a Tier2 centre). PMB Action List 15/10/01 Action List ----------- ONGOING... 1.11 PC & DB to determine which project management tools to be used 3.2 PC, NB, RB + EB to define experimental requirements 3.8 NG to provide details of Oversight Board timetable 4.1 TD to set up the PMB web pages 4.7 RPM to formulate guidance on costs of sending person to CERN on LTA. Some examples would be useful. Agree with Ken Peach and circulate to PMB (Les would use this information to inform potential applicants). 5.6 JCG to estimate manpower needs for Tier 1/A centre 5.7 PC to come up with a list of names for TB NEW... 6.1 PC to get TB documents onto appropriate web-site. 6.2 JG to circulate hardware specifications to EB. 6.3 ALL to prepare input on top-down approach to recruitment