Ragagnin Antonio

Abstract Reference: 30812
Identifier:
Presentation: Focus Demo
Key Theme: 1 Reduction and Analysis Algorithms for Large Databases and Vice-versa 

A web interface to federalize the outcome of large, cosmological, hydro-dynamical simulations

Authors:
Ragagnin Antonio, Dolag Klaus, Biffi Veronica, Cadolle Bel Marion, Hammer Nicolay, Krukau Aliaksei, Margarita Petkova, Steinborn Daniel, 

We constructed a full data center (http://c2papcosmosim1.srv.lrz.de/) for cosmological simulations (like www.magneticum.org) which allows to federalize the outcome of large, cosmological, hydro-dynamical simulations, which typically exceeds hundreds of terrabytes of plain simulation data. This service allows to deliver scientific data products of various levels to a broad scientific community. For our data portal, the full outcome of several, large, cosmological, hydro-dynamical simulations are stored on a HPC systems (like SuperMUC at LRZ in our case) and federalized via a web interface by running post processing on demand on a separate, much smaller scale computing system (like the C2PAP cluster in our case). Users are able to access data products via a user-friendly web interface, running on easy expandable, virtual server machines. Users browse through visualizations of cosmological structures, guided by queries on meta data to select galaxy clusters and galaxy groups of interest. Users interactively compose such complex, science oriented queries against a data base, where the meta data is stored. The meta data of the current queried objects in addition can be displayed or downloaded form of CSV tables for a more complex analysis by the user. For a finally selected object, different categories of data products are made available, where all of them will involve performing remote post processing on the original, raw simulation data. So far, we provide three main services: (I) SimCut allows the user to obtain a sub-volume around the selected object within the simulation, containing all the original, hydro-dynamical informations; (II)
Smac allows to obtain idealized 2D maps of various, physical quantities of the observed objects; (III) Phox allows perform virtual observations with various, existing and future X-ray telescopes. The large data produced by the cosmological hydro-dynamical simulations are organized to allow such post-processing to be performed in a fast and resource optimized way.