European Earth Science E-infrastructures Workshop Insights

Earth science e-infrastructures workshop Diego Scardaci, EGI.eu Technical Outreach Expert EGI OMB 29 January 2015 www.egi.eu www.egi.eu EGI-InSPIRE RI-261323 EGI-InSPIRE RI-261323

The workshop Brought together leading European e-infrastructures providers with Earth Sciences research communities e-Infrastructures EGI.eu The Geant Association EUDAT PRACE Research communities VERCE, CLIPC, EIDA, EPOS Solid Earth Sciences, GPS, British Geological Survey EGI-InSPIRE RI-261323 www.egi.eu

WS Agenda e-Infrastructures introduction Use cases presentations Three key technical session identified AAI (chair Lukas Hammerle, The Geant Association) Data management and computing (chair Giuseppe Fiameni, EUDAT) Cloud (chair Diego Scardaci, EGI.eu) Contents of each technical session chosen according to user needs Communities filled in a questionnaire before the WS EGI-InSPIRE RI-261323 www.egi.eu

Use cases VERCE (Alessandro Spinuso, KNMI) CLIPC (Maarten Plieger, KNMI) EIDA (Javier Quinteros, GFZ) EPOS (Daniele Balio, EPOS Consortium) Solid Earth Sciences (Horst Schwichtenberg, Fraunhofer SCAI) GPS (Francesco Casu, CNR-IREA) British Geological Survey (Simon Flower, British Geological Survey) EGI-InSPIRE RI-261323 www.egi.eu

VERCE Virtual Earthquake and Seismology Research Community in Europe VERCE platform: distributed computing and data management services portal for Earthquakes Simulation and Earth Model evaluation. VERCE Science Gateway: Based on gUSE/WS-PGRADE technology Workflows Globus GT5 and Unicore Data management layer Based on a federation of iRODS nodes which stores private and public data and models (GridFTP, iRODS microservices, mongoDB) Requirements AAI, Data manag. and sharing, information about resources available http://verce.eu EGI-InSPIRE RI-261323 www.egi.eu

CLIPC Climate4Impact portal Portal dedicated to the climate impact community: based on 21 use cases from e.g, Deltares, Alterra, UvA Data from Earth System Grid Federation (ESGF): distributed data repository with OpenID acess (6 PB of data), access is free AAI Portal login with ESGF open id identifier Uses x509 client certificates for server side data access Requirements Reduce security complexity (OpenID, MyProxy, SSL, ports, delegation issues, etc ) Data transfer is a bottleneck EGI-InSPIRE RI-261323 www.egi.eu

EIDA - European Integrated Data Archive EIDA is a distributed data center: securely archive seismic waveform data and related metadata, gathered by European research infrastructures provide transparent access to the archives by the geosciences research communities Currently around 350 TB from 130 networks and 4500 stations ArcLink protocol (GFZ technology) Member of EUDAT 2020 project Requirements Secure storage Users want to see the repository as a unique source of information (integration) AAI (transparent access) EGI-InSPIRE RI-261323 www.egi.eu

EIDA stations EGI-InSPIRE RI-261323 www.egi.eu

EPOS The European Plate Observing System (EPOS): multidisciplinary research is made possible for a better understanding of the Solid Earth Science processes long-term plan to facilitate integrated use of data, models and facilities from distributed and new research infrastructures Earthquakes, Volcanic Eruptions, Tsunamis, Tectonics, etc. Huge community: 244 RI, 138 Institutions, 22 countries, many sensors More than 2 PB of data (seismology + GNSS) EPOS will adopt an external ICT infrastructure Requirements AAI Data Integration Workflow engines, PID, data staging, move computation to data, processing EGI-InSPIRE RI-261323 www.egi.eu

Solid Earth Sciences part of EGI Earth Science community Embrace large variety of disciplines each having its own approaches and tools No general infrastructure for Earth Science Data from the climatology group stored via the Earth System Grid (ESG) on EGI resources Different AA mechanisms Requirements AAI Discovery and automatic download of data sets from computation jobs User interfaces via portals are not very popular in this discipline, different categories of users have to be catered for EGI-InSPIRE RI-261323 www.egi.eu

GPS Satellite Data Services: earth surface displacements maps geophysical parameters retrieval provided by 5 partners coming from different European countries each service needs an infrastructure EPOSAR service generation of displacement time-series by exploiting the existing huge SAR data use of external computational infrastructures can be considered Requirements processing of large data: from 150MB up to 10GB for 1 image according to the satellite algorithm to reduce the processing time WNs: 10-100 per processed area, RAM > 32 GB, Network >10 Gbps among WNs terabytes for each simulation, a solution is needed to stage data to computing EPOS Thematic Centre EGI-InSPIRE RI-261323 www.egi.eu

British Geological Survey British Geological Survey: Involved in many scientific areas (geophysics, etc..) owns an IT infrastructure Involved in EPOS Many web services already available Requirements Lack of expertise to scale with computing capabilities: need help to understand the potentiality of a distributed e-infrastructure Integration of scientific code into the infrastructure requires care EGI-InSPIRE RI-261323 www.egi.eu

AAI Session Issues X.509 is difficult for some users Group management is weak in eduGAIN/FIM Delegation, definition of roles Too many protocols in use How to continue? First step focus on definition of roles ? Collect use-cases: short-term actions and long-term actions. Trying to identify and define common roles specification for the Earth Science communities could be a use-case The e-infrastructure providers should be involved too in defining these roles Some use case requirements will be input for the G ANT 4 Enabling Users task AARC project EGI-InSPIRE RI-261323 www.egi.eu

Data and computing Session Complexity of software stack. Too many software layers may affect community services sustainability Deploy and operation in production of community services Reduce the number of pilots in favour of smaller-scale long-term collaborations Proceed gradually into the integration of e-Infrastructure services PID and DOI based on the same technology (Handle System) and serve two different utilization scenarios (data and publication) They can be combined (a DOI pointing to a publication which includes references to data sets through PIDs) but are not interchangeable RDA has a specific working group to discuss on these topics GEANT to contribute to EGI/EUDAT/PRACE calls for use cases to ensure networking needs and support for any AAI issues EGI-InSPIRE RI-261323 www.egi.eu

Cloud Session GEANT Cloud services is about providing access to existing cloud services in various ways Catalogue and compliance to small set of clear key requirements Procurement based on aggregated demand from NREN user base EGI offers a federated cloud for scientific communities Already supporting use cases coming from various scientific disciplines Ready to support more use cases It could be listed in the GEANT catalogue EGI-InSPIRE RI-261323 www.egi.eu

Conclusions & Follow-up GEANT, EGI, PRACE and EUDAT will collaborate on supporting scientific applications EGI/EUDAT/PRACE calls for use cases coming soon GEANT will contribute ensuring networking needs and support for any AAI issues e-Infrastructure services integration (e.g. EGI-EUDAT cookbook) 2 use cases identified to start pilots CLIPC Satellite Data / GPS Pilots setup will start in the next weeks EGI-InSPIRE RI-261323 www.egi.eu

Thank you www.egi.eu www.egi.eu EGI-InSPIRE RI-261323 EGI-InSPIRE RI-261323