Prev: P4.15 Next: P4.18

P4.16: Romelli, Erik
Erik Romelli (INAF - OATs)
Marco Frailis (INAF - OATs)
Samuele Galeotta (INAF - OATs)
Daniele Tavagnacco (INAF - OATs)
Davide Maino (University of Milano, INFN)
Claudio Vuerli, INAF-OATs Gianmarco Maggio, INAF-OATs Giuliano Taffoni, INAF-OATs


Theme: Data Science: Workflows Hardware Software Humanware
Title: Euclidizing external tools: an example from SDC-IT on how to handle software and humanware

Euclid is an upcoming space mission aimed at studying the dark Universe and understanding the nature of the so called Dark Matter and Dark Energy. The launch is scheduled at the moment for the 2021. A huge amount of data, up to more than 70 PB, will be produced by the two on-board instruments (VIS and NISP) and the data processing will be a crucial aspect of the mission, especially dealing with the performance of the code involved in the scientific analysis. Due to the expected amount of data and estimated number of cores a distributed computing system on several Science Data Centers (SDCs) has been implemented. To ensure a uniform environment in all SDCs, any software designed for Euclid must comply with a set of common rules and must be implemented in a predefined framework. Not all the code is designed and implemented ex novo for Euclid purposes; usually data analysis pipelines inherit already existing software tools, designed outside the Euclid Consortium. SDCs are in charge of the integration of external code within the official Euclid software environment. We will present an overview of how that was done at the Italian SDC in Trieste and present some practical examples related to the integration of an external tool into the Euclid environent. The topic will be discussed taking into account the technicality and focused on the crucial, but usually ignored, aspect of human interfaces.

Link to PDF (may not be available yet): P4-16.pdf