1 / 66

Dr. Isabel Campos Plasencia IFCA-CSIC

Dr. Isabel Campos Plasencia IFCA-CSIC. BUILDING AN EOSC IN PRACTICE. Cloud Services for Synchronisation and Sharing 28 - 30 January 2019, Roma. Outline. Background on EOSC The shortest possible version of the history

rozalia
Télécharger la présentation

Dr. Isabel Campos Plasencia IFCA-CSIC

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Dr. Isabel Campos Plasencia IFCA-CSIC BUILDING AN EOSC IN PRACTICE Cloud Services for Synchronisation and Sharing 28 - 30 January 2019, Roma

  2. Outline • Backgroundon EOSC • Theshortestpossibleversion of thehistory • Explaningtheconclusions of the EOSC HLEG: Minimum Viable Ecosystem • Whatis happening now? • A glimpseonthe status of EOSC-hubproject* • How can youjointheeffort: becomean EOSC serviceprovider* *Slidescourtesy of Tiziana Ferrari Isabel Campos 29° January 2019 / Rome

  3. European Commission Digital Single Market (DSM) • April 2016: European Cloud Initiative COM(2016) 178 • as part of the 'Digitizing Industry' package • EuroHPC: European Strategy on High Performance Computing • Widening access and building trust: Open Science • European Open Science Cloud: EOSCas the instrument to support Open Science Isabel Campos 29° January 2019 / Rome

  4. Idea and Realization of an EOSC The history (The shortest possible version) April 2016: European Cloud Initiative COM(2016) 178 as part of the 'Digitizing Industry' package The EOSC aims to give Europe a global lead in scientific data infrastructures…. It will offer 1.7 million European researchers and 70 million professional in science and technology a virtual environment with free at the point of use, open and seamless services for storage, management, analysis, and reuse of research data, across borders and scientific disciplines. Source : COM(2016) 178final, p.6 Isabel Campos 29° January 2019 / Rome

  5. Policystepstowards EOSC implementation: • 23.11.2018 • EOSC Launch Event • Vienna Declaration on EOSC emphasizing the need to ensure smooth and successful implementation: • https://eosc-launch.eu/fileadmin/user_upload/k_eosc_launch/EOSC_Vienna_Declaration_2018.pdf) • Appointment of a • Governance Board (Member States) • Executive Board (EC appointed) 14.03.2018 Implementation Roadmap After consultation with the EU Member States, the EC the adopts an Implementation Roadmap for EOSC: See details yourself at: https://ec.europa.eu/research/openscience/pdf/swd_2018_83_f1_staff_working_paper_en.pdf June 2017 June 2018 November 2018 March 2018 07.11.2017 1st EOSC Summit The Commission made the EOSC Declaration available to all scientific stakeholders for them to provide their endorsement and commitments to the realisation of the EOSC by 2020. 11.06.2018 2nd EOSC Summit brought together key players for the implementation of the EOSC, representing scientific fields, research funding organizations and officials from ministries of Member States and Associate Countries. Isabel Campos 29° January 2019 / Rome

  6. EC High LevelExpertGroups: 1st and 2nd 1st 2nd • Provide methodological advice to the Commission on the fulfillment of the EOSC objectives • Care should be taken that EOSC does not duplicate existing initiatives and, adds value by integrating existing initiatives as well as the outcome of past, ongoing and forthcoming projects. …to provide strategic advice to the Commission on the strategy for the European Open Science Cloud initiative as part of the Digital Single Market. June 2016 June 2017 December 2018 Isabel Campos 29° January 2019 / Rome

  7. Role of the 2nd HLEG EOSC 2nd High Level Expert Group (Jun 17 – Dec 19) • Aim: To mark a transition towards the practical implementation of the EOSC and to set the scene to the practical launch the EOSC • Focus: Governance Structure, Rules of Participation and Business model options • 32 recommendations are provided, clustered in Implementation, Engagement and Steering Isabel Campos 29° January 2019 / Rome

  8. Key elements in thereport • Vision – The EOSC as interlinking people, data, services & trainings, publications, projects, and organisations across borders and scientific disciplines • MVE – Discussion on how to make EOSC a Minimum Viable Ecosystem • Business models – 3 valid alternatives for funding the EOSC have been outlined • The set of practical recommendations – including on EOSC portal Have a look yourself: https://publications.europa.eu/en/publication-detail/-/publication/5253a1af-ee10-11e8-b690-01aa75ed71a1/language-en/format-PDF/source-80622260 Isabel Campos 29° January 2019 / Rome

  9. Some initialconsiderations …EOSC can only be graduallyimplemented. …EOSC will NOT provide science with functionalitiesthat are not already provided by large Cloud providers for generic purpose. However, it has the potential to deploy added value services for science: • .... for providing seamless access to instruments and the borderless exchange of data and knowledge (FAIR) and to produce results beyond the sum of the individual findings. • ... for simplifying funding models and moving away from the project based funding nightmare. • ....for decoupling scientific domain knowledge from IT domain knowledge experts and let each of them focus on their area. • Essentially now, dedicating too much time to IT “technical” duties ends your career as a domain scientist. Isabel Campos 29° January 2019 / Rome

  10. As a High Level Objective, EOSC aims at allow for universal access to data, and become a new level playing field for EU researchers Researcher • Easy access through a universal access point for ALL European researchers • Cross-disciplinary access to data and services unleashes potential of interdisciplinary research • Services and data are interoperable (FAIR data) • Data funded with public money is in principle open (as open as possible, as closed as necessary) • EOSC will help increase recognition data intensive research and data science • Access to all European research data • Access to world-class data services • Clear rules of use and service provision • FAIR data tools, training and standards CERN, EMBL, ELIXIR, etc. Institutional repository Member State Infrastructure New provider/ service Seamless environment enabling interdisciplinary research Source: RTD

  11. The EOSC will allow for universal access to data and a new level playing field for EU researchers Researcher • Easy access through a universal access point for ALL European researchers • Cross-disciplinary access to data and services unleashes potential of interdisciplinary research • Services and data are interoperable (FAIR data) • Data funded with public money is in principle open (as open as possible, as closed as necessary) • EOSC will help increase recognition data intensive research and data science Likelytheroot of theproblemis thatknowledgeremainsconfinedinside disciplinary silos • Access to all European research data • Access to world-class data services • Clear rules of use and service provision • FAIR data tools, training and standards CERN, EMBL, ELIXIR, etc. Institutional repository Member State Infrastructure New provider/ service Seamless environment enabling interdisciplinary research Source: RTD

  12. EU RI & eInfras EOSC COMMERCIAL SERVICES • ESFRI OPEN DATA PORTALS EURO-HPC OTHERS… INDUSTRY RESEARCH PUBLIC SECTOR PUBLIC INSTITUTIONS OTHER OPEN DATA PORTALS SMEs PPP

  13. Buildingthe EOSC in practice Startingwith a MinimumValueEcosystem

  14. EOSC as a Minimum Viable Ecosystem Policies TechnicalConditions Human Conditions Interoperable Services Open Data Software Developers Researchers Infrastructure Managers EOSC aims to put Europe at a global lead regarding scientific data infrastructures. The provision of infrastructures, technology development, and human resources to support it will take place in a very heterogeneous landscape. Addressing this challenge requires the definition of a smallest common denominator: the EOSC Minimum Viable Ecosystem (MVE).

  15. The way to do it is through MVP strategy • Minimum Viable Product is • “a version of a new product which allows a team to collect the maximum amount of validated learnings about customers with the least effort.” • MVP includes • “the feature or features required to solve a core problem for a set of users and be released to market”. Isabel Campos 29° January 2019 / Rome

  16. EOSC as a developing ecosystem • Minimum Viable Ecosystem would enable EOSC to emerge as • a collaborative effort, • in an iterative way: EOSC is not a project, is an iterative process • Identification and Understanding of the needs • Finding the opportunities which includes • identification of users, • their actions and desired results as well as • identification of pains and gains of each action. • Deciding what features to build – • what pains to remove with product that • gives gains to the user. Recognise that the implementation of the European Open Science Cloud is a process, not a project, by its nature iterative and based on constant learning and mutual alignment(EOSC Vienna Declaration, 23.11.2018) Isabel Campos 29° January 2019 / Rome

  17. There are three+one key user groups • Researchers • they are the customer and customer is king • Software developers • they are the makers of the whole thing • they make the ecosystem rich or poor • Infrastructure providers • they get an opportunity for new businesses on their platforms • Research funding organizations • they foot the bill • they care for efficiencies of the system Isabel Campos 29° January 2019 / Rome

  18. Isabel Campos 29° January 2019 / Rome

  19. An MVE will emerge ifpolitical, technical and human/sociological conditions are met Policies TechnicalConditions Human Conditions Interoperable Services Open Data Researchers Software Developers Infrastructure Managers Isabel Campos 29° January 2019 / Rome

  20. Realityis more complex/richer: inter-dependencies Policies TechnicalConditions Human Conditions Researchers Software Developers Infrastructure Managers Interoperable Services Open Data Isabel Campos 29° January 2019 / Rome

  21. What can EOSC do forResearchers Call for the European Open Science Cloud to provide all researchers in Europe with seamless access to an open-by-default, efficient and cross-disciplinary environment for storing, accessing, reusing and processing research data supported by FAIR data principles (EOSC Vienna Declaration, 23.11.2018) Confirmingthat the vision of the European Open Science Cloud is that of a research data commons, inclusive of all disciplines and Member States, sustainable in the long-term: Isabel Campos 29° January 2019 / Rome

  22. Open Science Commons:When implemented… Researchers from all disciplines will have easy, integrated and open access to the advanced digital services, scientific instruments, data, knowledge and expertise they need to collaborate and achieve excellence in science, research and innovation. They feel engaged in governing, managing and preserving these resources for everyone’s benefit, with the support of all stakeholders.

  23. Scientific infrastructures produce large amounts data (PBytes…): Factual European researchers face data fragmentation and unequal access to quality information sets Researcher • Access to data and resources CERN, EMBL, ELIXIR, etc. Institutional repository Member State Infrastructure Limited and limiting access for an ordinary European researcher Source: RTD

  24. Scientific infrastructures produce large amounts data (PBs…): Factual European researchers face data fragmentation and unequal access to quality information sets • Fragmented access (across scientific domains, countries and governance models; varying access policies) • Limited cross-disciplinary access to data sets (i.e. interdisciplinary research) • Non-interoperable services and data • Closed data Researcher ? • Access to data and resources CERN, EMBL, ELIXIR, etc. Institutional repository Member State Infrastructure Limited and limiting access for an ordinary European researcher Source: RTD

  25. Scientific infrastructures produce large amounts data (PBs…); Factual Let us answer out of own experience • Fragmented access (across scientific domains, countries and governance models; varying access policies)? • Large Collaborations have a strategy to solve this problem for themselves: dedicated funding lines • For the average research group, there are nightmare scenarios. • Limited cross-disciplinary access to data sets (i.e. interdisciplinary research): Factual • Is it needed? Sometimes it is. • Difficult to overcome, as today´s scientific challenges and career evaluation policies impose a huge specialization on researchers. • Non-interoperable services and data: Factual • Closed data: Plenty, and will continue so (embargo periods) Limited and limiting access for an ordinary European researcher Source: RTD

  26. Challenge: Bringingeveryoneonboard Top EOSC needs to address the problem that researchers in Europe have insufficient access to e-Infrastructures. Up Down Bottom Policymaking ScientificResearch • Top of thewish-listforResearchers: • Flexibleaccessto Computing Resourcesforeverydaywork; • Store Data, in a self-controllableway, in largeshared pools; • Real accesstoFast Networks

  27. Bringingeveryoneonboard • The adoption can only be incremental, pilotedinitiallybylargecommunities • Howtoprojectan EOSC for alltheresearchers in Europe? • BTW: are open sourcesync & share solutionsready ? • Coordinated (potentially “federable”) country-levelsolution for the so calllongtail of users • Commonresearchprojects are done in teams of between 5-15 persons. • Ex. In myenvironment, Postdocs and Phdstudentsspend a substantialamount of time figuringout, how and where, tomove a few Terabytes of Lattice QCD configurationsfrom A to B. Isabel Campos 29° January 2019 / Rome

  28. SomePolicyRecommendationstoIncentivizeResearcherstoadopt EOSC P5 - Improved Fast Network access to EOSC data through policies for better harmonisation of research networks with particular attention to “last mile” – researchers have to be able to get to the data P1 - Ensure that researchers who lodge curated data in trusted repositories are recognised in funding and promotion schemes P4 - Ensure policy initiatives focused on strengthening research infrastructure provider participation in EOSC – both at EU level and member state level  Isabel Campos 29° January 2019 / Rome

  29. SomePolicyRecommendationstoIncentivizeResearcherstoadopt EOSC P1 - Ensure that researchers who lodge curated data in trusted repositories are recognised in funding and promotion schemes • A domainscientistdedicatetothismatters, would do a veryusefulworkperhaps, butwould ruin hiscareer. • Thecommunityneedstoagreeoncertain rules tostore data and itsmetadatarelatedinformation • IT supportisneededtoactuallymaintainthoserepositories: technicianlevel. Isabel Campos 29° January 2019 / Rome

  30. Infrastructure Managers Recall that the Council - in its conclusions of 29 May 2018 - welcomed the implementation roadmap and the federated model for the European Open Science Cloud. Resolveto harness the many ongoing and planned activities at EU and Member States level to cooperate in establishing an inclusive partnership with a view to developing the European Open Science Cloud as a federated infrastructure that can enhance value-based, open, trusted, user-centric digital services across borders within the Digital Single Market (DSM) (EOSC Vienna Declaration, 23.11.2018 ) Isabel Campos 29° January 2019 / Rome

  31. From a technicalstandpoint:Weknowhowtofederateinfrastructures InfrastructurebuildingCapacity: Communityfunded and/orNationalInfrastructures 2001 2004 2010 2017 BuildingCapabilitiesviaUserCommunityEngagement

  32. TechnicalStandpoint in FederatingInfrastructuresand itworks (seeaccounting.egi.eu ) : 2001 2004 2010 2017 CPU Wall Time (Billionhours) provided

  33. +20% utilization of computing in 2018 TechnicalStandpoint in FederatingInfrastructuresand itworks (seeaccounting.egi.eu ) : 2001 2004 2010 2017 CPU Wall Time (Billionhours) provided

  34. TechnicalStandpoint in FederatingInfrastructuresand itworks (seeaccounting.egi.eu ) : 2001 2004 2010 2017 • 2018 Figures • 4.4 Billion CPU core wall time delivered in 2018 = 220 M€ in Amazon • 970,600 computing cores • 356 PB disk & 380 PB tape storage • 1170 Open Access publications • +41 new international projects • 31 large scale ESFRI projects/landmarks supported Isabel Campos 29° January 2019 / Rome

  35. TechnicalStandpoint in FederatingInfrastructures: Makingtheleapto EOSC 2001 2004 2010 2017 Buildingonwhatwehave Integration at the AAI level (the Virtual Organizationmodelworked, butit has limitations) Monitoring Grey Zone EGI FederatedModel Open Science EOSC Model Automatization of Resourceprovisioning (italwaysworkedmanually: VO configuration) Integration at thesecuritylevel Technical and Policy - wise • In a nutshell • PolicyLevel: Actionsto be supportedbytechnicaldevelopments, notably in thefront of Security (data privacypolicies) and accesspolicies • TechnicalIntegrationLevel: workto be done/extended, buildingonthepreviousexperience.

  36. Incentives forResearchInfrastructure Managers“GeneratingEconomicGrowthviaaddedvalue” • Integration resources in the EOSC is a way to achieve a higher a more efficient usage. • Supporting International Research Collaborations • Prototyping and piloting Innovative Services for cutting-edge Research Economy of Scale Expandinguser base EarlyCareerProfessionalsEducation Independent IT Research at theState of the Art Isabel Campos 29° January 2019 / Rome

  37. PolicyRecomendationstoIncentivizeInfrastructure Managers • P3- Discuss the practical elements of the business model with Member States P5 – Incentivize Infrastructure managers through funding streams & maintenance costs for shared facilities and data repositories. Isabel Campos 29° January 2019 / Rome

  38. EOSC as ecosystem for Software Developers Highlightingthat Europe is well placed to take a global leadership position in the development and application of cloud services for Science. Reaffirmthe potential of the European Open Science Cloud to enable first-class data-driven science and to stimulate new business models benefiting our society and the economy. Recognise that such services will create opportunities for both public and private sectors, notablyby intensifying reuse of public sector information while preserving data integrity, and ensuring access, transparency within and across borders. (EOSC Vienna Declaration, 23.11.2018) Isabel Campos 29° January 2019 / Rome

  39. Incentives for Software Developers: “Engaging Human Talent” Breakthrough ideas leading to innovation need to be awarded with the proper Recognition • Indicators of recognition in popular software repositories (eg. Github) • # of forks • # of starsawardedbypeers and users • # of downloads • AdditionalIndicatorsthroughEOSC • # of servicesbuiltusing a software component • #of deploymentsonResource Centers • # of users/scientificcommunitiesserved

  40. Incentives for Software Developers: “Engaging Human Talent” • “EOSC-Ready” as a Branding for software products, will harness the potential of the European developers, both in academia and industry. • What does it mean “EOSC-Ready” ?Trustable • Open Source  contributedupstream (potentiallysustainable) • Based on Open Standards • Adheres to code style quality standards Software Quality • It respects the EC directives on cloud security (NIS), personal data protection (GDPR) and Free Flow of non-personal Data (FFD). EOSC - Ready

  41. Incentives for Software Developers: “Engaging Human Talent” “EOSC-Ready” as a branding for software products, will harness the potential of the European developers, both in academia and industry. The Devil might be in the details: how & who would be awarding such label? Nothing is black or white, there is a grey scale: • Is Closed Source software not good to be used to deploy services in EOSC? • Certainlynotfor AAI. • Norforcoreservicesthatsupportthefederationlayer(IMHO) • Open Standards: difficult compromises always among which standard… • Software Quality procedures: need to substantiated in an agreable way (building on community recognized best practices) • EU Directives: Is the EC Market too regulated? How will this affect EOSC? (interesting debate)

  42. RecommendationstoIncentivize Software Developers P2 – Stronger Open policies for trusted software and associated data services Trusted services rely on the existence of a software ecosystem adhering to the principles of software openness, which builds on open standards. The recommendation isthepromotion of an open software ecosystem, supported by a consistent, community-recognized, software certification*. (*) i.e. NOT an ISO-likecertification

  43. Business modelfor EOSC Commit to support service provision for the European Open Science Cloud by helping connecting relevant national and disciplinary nodes to the pan-European level. Reaffirm the potential of the European Open Science Cloud to enable first-class data-driven science and to stimulate new business models benefiting our society and the economy. (EOSC Vienna Declaration, 23.11.2018)

  44. Business models: accessingmodes ExcellenceDrivenAcces Exclusively dependent on the scientific excellence, evaluated through peer review conducted by experts • Agreement between the User and the e-Infrastructure that will lead to a “payment” for the Access MarketDrivenAcces • Private providers will envision a fee-model for the Access Wide Acces Guarantees the broadest possible Access to scientific data and digital services to Users wherever they are based

  45. Business models: accessingmodes ExcellenceDrivenAcces Exclusively dependent on the scientific excellence, evaluated through peer review conducted by experts In practical terms A model based on Wide Access mode, modulated by a negotiated Agreeable Access restriction, is the pragmatic way to start moving with EOSC • Agreement between the User and the e-Infrastructure that will lead to a “payment” for the Access MarketDrivenAcces • Private providers will envision a fee-model for the Access Wide Acces Guarantees the broadest possible Access to scientific data and digital services to Users wherever they are based

  46. Business models: Someof theFinancing Instruments underanalysis Provideresourcestodevelopinnovativeserviceswhenthey are not economicallyprofitable in commercialproviders Directfunding Capacitymismatchings and accessproblemsforoutsidestakeholder Cloud vouchers (“coins”) • EasywayforCommercialproviderstoaccesstheMarketplace • Poor modeltomaintain data sets in view of themoney-flow gaps in • Researchfunding, and also data curation/preservationlifetimes. Hybridapproaches Supports access by both commercial and non-commercial organizations into market, allowing both groups to work in areas where they have specialized abilities.

  47. Inmediate challenges Challenge-2 Cross-border and cross-discipline research collaboration EOSC should provide a collaboration mechanism for research groups that respects the respective settings for security and privacy. Depending on the needed resources for computing, the system also needs to provide ad-hoc resources for intensive computations. Rulesfor collaboration shall include the potential of building dynamic ad-hoc groups. Challenge-1 Enable integration of and access to resources that will be federated in the EOSC Follow an open interface approach, which permits a seamless integration/adaptation to existing Hubs. Based on open standards APIs and protocols. Rules shall be set to build services based on open standard APIs and protocols for resource and service portability reasons.

  48. EC supporttotheimplementation of EOSC A short summary

  49. Call INFRAEOSC-04-2018-2019 ALLOCATED: several Connecting ESFRI infrastructures through Cluster projects (95M€) Call INFRAEOSC-01-2018-2019 ALLOCATED: OCRE Access to Commercial Services through the EOSC-hub 12M€ Call INFRAEOSC-05-2018-2019 TBD Setup of an EOSC coordination structure(10M€) Coordination of EOSC-relevant National Initiatives and support to prospective EOSC service providers (30M€) (c) FAIR data uptake (10M€) Call INFRAEOSC-02-2018-2019 TBD Development of Innovative Services (28,5M€) Call INFRAEOSC-06-2018-2019 TBD Development of the EOSC-Portal (2M€) April 2019 December 2020 January 2017 January 2018

  50. EOSC-hub expected impact • The major impact of the EOSC-hub project will be a significant reduction in the fragmentation of the IT facilities, services and tools for data-intensive research and innovation in Europe.

More Related