Test & Improvement of a Metadata-Crawler for HPC Simulations

Technische Universität München
December 18, 2022
Offerd Salary:Negotiation
Working address:N/A
Contract Type:Other
Working Time:Negotigation
Working type:N/A
Job Ref.:N/A
Test & Improvement of a Metadata-Crawler for HPC Simulations

https: // www. castordoc.com/blog/what-is-metadata

14.11.2022, Studentische Hilfskräfte, Praktikantenstellen, Studienarbeiten

as Student Assistant (HiWi) / Research Internship / Study Project / Thesis for TUM-students within Informatics, Data Engineering, Mechanical Engineering, Engineering Science, Electrical Engineering, Aerospace etc.

We are working on solutions to make data from High Performance Computing (HPC) findable and reusable by other research groups. This is supported by attaching rich metadata to the research data. The relevant metada-ta fields are described in ontologies. Filling these ontologies by hand, however, is infeasible. Therefore, in this work, the metadata extraction is to be automated using a python based script. The tool should first extract metadata form runs of our in-house computational fluid dynamics (CFD) code and then successively extend to data from the the HPC system used and post-processing algorithms. Therein, flexibility to work on inputs from other sources, e.g. other codes or experimental data is to be ensured.

Tasks • Familiarization with CFD and HPC workflows • Familiarization with metadata-ontologies • Improvement of a (existing) python toolkit to extract metadata

Requirements • Knowledge of python • Knowledge of Linux-CL

Benefits • Flexible working hours • Joint-mentoring by TUM and LRZ: exclusive experience with HPC-clusters • Experience with state-of-the art metadata concepts • Insight into a nationwide research project • Credentials of prestigious research facilities (TUM and LRZ) • Recognition within your study program (just for thesis / internships / projects) • Payment will be according to TV-L

Contact Benjamin Farnbacher [email protected] 089.289.16094

Kontakt: [email protected]

Mehr Information

Advert Advert as PDF, (Type: application/pdf, Größe: 92.9 kB) Datei speichern

From this employer

Recent blogs

Recent news