Mainz 2014 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
T: Fachverband Teilchenphysik
T 35: GRID-Computing
T 35.5: Vortrag
Montag, 24. März 2014, 17:45–18:00, P15
Dynamic data allocation and long term storage access in the German ATLAS cloud — •Thomas Hartmann1, Gen Kawamura2, Kai Leffhalm3, Friederike Nowak3, Günter Duckeck4, Torsten Harenberg5, Marisa Sandhoff5, Rodney Walker4, Michael Böhler6, Anton Gamel6, Andrzej Olszewski7, and Jan Erik Sundermann6 — 1KIT Karlsruhe — 2Universität Göttingen — 3DESY Hamburg — 4LMU München — 5Bergische Universität Wuppertal — 6Universität Freiburg — 7Instytut Fizyki Jadrowej Krakow
The global data processing and simulation of HEP data is a comparable effort as the actual data taking by the LHC experiments as ATLAS. For a successful analysis of the recorded events, several different Monte Carlo simulations have to be performed. Like the event data, the computing resources are distributed globally. Organizing and making these resources available is the task of the Worldwide LHC Computing Grid. Within the Grid, Gridka is the local organization of computing and storage resources at computing tiers in Austria, the Czech Republic, Germany, Poland, and Switzerland. Within Gridka, Tier-2 centres provide computing resources for user analyses and MC production and, as one of the larger Tier-1 centres, event data and computing resources are provided at KIT Karlsruhe. We present the cloud's implementation and experience of ATLAS' approach for dynamic data allocation ("federated access") in addition to the classic hierarchic approach. Furthermore, the talk will focus on the cloud's development of resources for long term storage access for end users.