Type:
Personnel de support
Type de contrat:
CDI
Lieu de travail:
Belval
Contexte
Under the responsibility of the Head of Information Systems, you will join a team of two persons in the HPC Service to help the operation, the administration and the maintenance of the LIST’s in-house HPC system, the Visualization Wall cluster and the Big Data-oriented cluster.
These three facilities support LIST researchers in solving advanced and complex computation problems in environmental, biological, material sciences and technologies as well as in big data analytics.
Description
Your primary responsibility is the system administration of the three cluster systems, in this framework, you will be in charge of the:
- Administration and maintenance Linux servers (CentOS / RHEL based)
- OS installation and configuration
- Adapting the HPC scheduler (SLURM) depending on user’s needs
- Managing the backups and archival
- System planning and extension and involvement in the acquisition of new components
- Users support (software installations, etc.)
- Corrective, progressive, preventive and maintenance of the HPC system (hardware, software)
- Processing and analysis of anomalies (hardware, software)
- Understanding and analysis of users’ needs to better configure and use the HPC / Vizwall / Big Data system
- Estimation and planning of the tasks
- Subcontracted work supervision
- Drafting and maintenance of the documentation
- Participation in internal meetings and in European consortia and fora (ETP4HPC, PRACE, Terratec, etc.)
In addition, you may be involved in scientific programming: tasks include supportive work on the implementation, maintenance and upgrade of complex numerical models (development, experiment design, simulations). You will support researchers in optimizing their scientific code (performance debugging).
Profil
Education
- Master Degree in Computer Science or other scientific domain with a professional experience (minimum 4 years) in HPC operation
Competencies
- Demonstrated experience in HPC operation and management
- Good understanding of HPC programming technologies (MPI, openMP, TBB, Fortran, etc.)
- Good knowledge of cluster schedulers (SLURM / YARN)
- Strong experience in programming and scripting
- Good understanding of various compilers (Intel, GCC toolchain, PGI, etc.)
- Very good knowledge of HPC related hardware technologies (Infiniband / Intel OPA, AMD/Intel CPU arch, etc.)
- Very good knowledge of Linux
- Knowledge of the tools, libraries and formats like NCO, numpy, MPICH, PBS, NetCDF, CRON, GDAL, IDL/ENVI, ErdasImagine, ESRI ArcGIS is a plus
- Open-minded, flexible and think interdisciplinary
- Creative, autonomous, proven organizational and communicational skills
- Organized and rigorous, sense of responsibility
- Result and service-oriented
Language
- Proficiency in English both spoken and written is mandatory
- French speaking is a plus
Deadline for application: 29 February 2020
