EU 2010 Problems Opportunities HEP VRC

EU 2010 – Problems & Opportunities (HEP VRC et al) IT Programme of Work Meeting Morges, November 2009

Introduction • In this talk I’d like to stay high-level – some details are included but an in-depth discussion of these would mean that we risk to miss the “big picture” Which is: • If used optimally we could: Serve (well) our users & help exploit LHC to the max; Strengthen collaboration with related disciplines (partners in the various projects); Justify future – virtually indefinite (but definitely finite) – funding.

Overview • This talk will not review the status of the HEP VRC, nor that of EGI In. SPIRE SA 3 in detail – Slides covering these topics are attached to the agenda fyi – as well as the “final” proposals… • How the effort is integrated into IT is of course important (see Frederic’s IT 2010? ) but is not the subject of this talk • It will focus on the (IMHO) exciting prospects that funding in this area opens up

HEP VRC in a nutshell • Actually (part of) two proposals: – EGI In. SPIRE – Services for user communities that are heavy users of DCIs and have multi-national dimension (often referred to as “HUCs”) – ROSCOE – one of several proposals for a (set of) Virtual Research Communities: Support activities • If reasonably funded, these projects should allows us to: – Manage transition from EGEE & other funding to “EGI” – Contribute to the successful exploitation of the LHC VRC = Virtual Research Community = EU speak for 1. 2. 3

MANPOWER – SKIP AT THIS STAGE!

PROBLEMS

Executive Summary • The problems [ described next ] are: Few in number; Rather well understood in scope; Surmountable • The opportunities are significant for: Ø CERN IT and our contribution to LHC exploitation; Ø Collaboration with related disciplines and hence Ø Justifying future funding. • The scale of such funding is miniscule wrt potential benefit

Problems & Resolution 1. Some staff/fellows have already left – others have contracts terminating end 2009 • ü Size of the [ remaining ] problem is 2 people for period up to start of EGI & related projects – some funding is available via INFN; more via experiments? Others? The 2 people involved have contracts at IT-T 1 2. The funding that we [ hope to ] get from the EU might be only sufficient to fund fellows and maybe < 36 months We need a reasonable ratio of staff / fellows (retention of expertise, supervision load etc. ) • IT / PH / experiment / “partner” funds could be used to “upgrade” fellows to CERN LDs are required J “Buy-in” from experiments would be valuable - and the “gap” is between € 93 K / € 100 K & (full) LD costs • Problems solved [ in principle ] – let’s move on…

OPPORTUNITIES

Introduction • To motivate this section I’d like to refer to two Microsoft Research workshops that I was invited to in Venice, Italy [ weekends ] • The 2 nd had two [ noble ] themes: one of which was education MIf you mess up in education it can take decades to discover and more decades to “correct” – cost / impact “immeasurable” Ø Timescale: comparable to that of the (s)LHC

Problems – again • I am not trying to gloss over the fact that this will involve work – including some that cannot be considered strictly mainstream [ other disciplines? ] • But I believe that it is not only consistent with our mission but also adds significant value to the org. • How do you measure the value? – What has been the cost – and benefit – of the CERN Program Library? • Actually a small # of FTEs compared to what we spend these days but enormous direct & indirect value – What was the cost and what has been the value to the world of Sir Timothy’s “baby”

Opportunities – again • I am not trying to gloss over the fact that this will involve work – including some that cannot be considered strictly mainstream [ other disciplines? ] • But I believe that it is not only consistent with our mission but also adds significant value to the org. • How do you measure the value? – What has been the cost – and benefit – of the CERN Program Library? • Actually a small # of FTEs compared to what we spend these days but enormous direct & indirect value – What was the cost and what has been the value to the world of Sir Timothy’s “baby”

And? • We will be required to work with [ and for ] other communities in the EGI Service & Support projects – Take the money & run or try for some +ve feedback for CERN / HEP? • We have the – real IMHO – opportunity to a) Provide significant long-term socio-economic value – measurable over decades & 108+ people b) Influence future funding (go back to a)

In Practice? L In reality we have little choice but to continue down the road we are already on • We can loath it or leverage it • I hope that it is clear what my recommendation would be… • What does this mean for IT? – It re-enforces an already visible world-class role – It strengthens our role wrt experiments and inside the lab: IT for the major physics lab is not IT for Migros… Ø My internal metric is to change “machine – experiment” partnership to something with at least a hint of an extra dimension…

What do I have to do? • Evangelise – and convince others to too! – Experiments, management, directorate, … • Do not agree with the pessimists…

What do I get? • The potential is for further funding (global IT manpower) and increased recognition • Something noble…

Experiment Support (HUC, SSC, experiment support, . . . ) Project WP / Task Activity EGI In. SPIRE SA 3 / TSA 3. 2. 1 Dashboards – HUC support EGI In. SPIRE SA 3 / TSA 3. 2. 2 Ganga / Diane – HUC support EGI In. SPIRE SA 3 / TSA 3. 3 VO Services – ATLAS DDM, Panda; CMS Ph. EDEx, Fro. NTier, DBS, T 0 AST, Site. DB, CRAB; ALICE Alie. N, LHCb DIRAC, … [ Largest injection of effort & most important part of project to cope with demands of LHC startup and early data taking ] ROSCOE NA 2/3 Coordination, Dissemination, Training (unfunded) ROSCOE SA 1 User Support: Integration & Operations support “aka EIS” [ increased focus on operations ] ROSCOE SA 2 “Gateways” – synergies with Dashboards ROSCOE SA 3 Targeted application porting & Distributed Analysis Support From experiment input to “Activity SAn: Services for Communities of Heavy Grid Users”

POSTSCRIPT

IBM Dictum • There are no such thing as problems – only opportunities • Sometimes there are unsurmountable opportunities

MANPOWER – DETAILS

HEP Manpower Summary Partner WP Effort Goals BHAM SA 3 1 FTE Distributed Analysis Support – focus on ATLAS CERN NA 2 NA 3 SA 1 SA 2 SA 3 0. 5 FTE 2 FTEs Liaison (un-funded) Dissemination & Training (un-funded) User & Operations support Dashboards (“portals”) Distributed Analysis Support, Integration Support CESNET SA 1 1 FTE Support for Distributed Analysis DESY SA 1 SA 3 1 FTE Operational support for ILC Targetted application porting for ILC GSI SA 1 2 FTEs Operational support for FAIR IC SA 3 1 FTE Distributed Analysis Support – focus on LHCb INFN SA 1 SA 3 2 FTEs User & Operations support Integration support UIO NA 2 SA 1 0. 5 FTE 1 FTE Liaison to EMI Distributed Analysis Support – focus on ATLAS Based on real costs for real people with overheads – except CERN: € 100 K/FTE year

Distributed funding for Distributed Analysis Support – a Sustainable Model? HEP Manpower – at CERN Partner WP Effort Goals BHAM SA 3 1 FTE Distributed Analysis Support – focus on ATLAS CERN NA 2 NA 3 SA 1 SA 2 SA 3 0. 5 FTE 2 FTEs Liaison (un-funded) Dissemination & Training (un-funded) User & Operations support Dashboards (“portals”) = 3 funded fellows / junior staff Distributed Analysis Support, Integration Support CESNET SA 1 1 FTE Support for Distributed Analysis DESY SA 1 SA 3 1 FTE Operational support for ILC Targetted application porting for ILC GSI SA 1 2 FTEs Operational support for FAIR IC SA 3 1 FTE Distributed Analysis Support – focus on LHCb INFN SA 1 SA 3 2 FTEs User & Operations support Integration support UIO NA 2 SA 1 0. 5 FTE 1 FTE Liaison to EMI Distributed Analysis Support – focus on ATLAS = 4 “INFN” funded fellows? Based on real costs for real people with overheads – except CERN: € 100 K/FTE year

ROSCOE – CERN Budget

In. SPIRE SA 3 “HEP” Manpower Year 1 Year 2 Year 3 Year 4 TSA 3. 1 6 PM 6 PM Activity Management TSA 3. 2. 1 24 PM - Dashboards TSA 3. 2. 2 24 PM - Ganga + Diane TSA 3. 3* 108 PM 102 PM - VO services TOTAL 486 PM * 24 PM allocated to INFN in years 1 – 3 “with the proviso that these people will get INFN money for staying at CERN virtually full time” [ and assuming no further cuts / surprises! ] Numbers to be confirmed! Based on EGI discussions 10 -NOV-2009