KERDATA Research team
Scalable Storage for Clouds and Beyond
- Leader : Gabriel Antoniu
- Type : Project team
- Research center(s) : Rennes
- Field : Networks, Systems and Services, Distributed Computing
- Theme : Distributed and High Performance Computing
- Partner(s) : Institut national des sciences appliquées de Rennes,Université Rennes 1,École normale supérieure de Rennes
- Collaborator(s) : Institut de recherche en informatique et systèmes aléatoires (IRISA) (UMR6074)
The KerData project-team is namely focusing on designing innovative architectures and systems for scalable data storage and processing. We target two types of infrastructures: clouds and post-Petascale high-performance supercomputers, according to the current needs and requirements of data-intensive applications.
Examples of such applications are:
- Cloud data analytics applications (e.g., based on the MapReduce paradigm) handling massive data distributed at a large scale.
- Advanced (e.g., concurrency-optimized, versioning-oriented) cloud services both for user-level data storage.
- Large-scale simulation applications for Exascale supercomputers.
Convergence of Extreme-Scale Computing and Big Data Infrastructures
- High-performance storage for concurrent Big Data applications
- Big Data analytics on Exascale HPC machines.
Advanced data processing on Clouds
- Optimizing MapReduce-based data-intensive processing.
- Stream-oriented, Big Data processing on clouds.
- Geographically distributed workflows on multi-site clouds.
I/O management, in situ visualization and analysis on HPC systems at extreme scales
- Scalable I/O and in situ visualization of HPC simulations on post-Petascale platforms using dedicated cores.
- Mitigating I/O interference in concurrent HPC applications through the investigation of cross-application interference and I/O prediction.
- Optimized architectures for in situ visualization and advanced processing.
International and industrial relations
- NCSA/UIUC ANL: active collaboration with JLESC (Urbana-Champaign) on concurrency-optimized I/O for post-Petascale infrastructures
- BigStorage: a Marie Curie Initial Training Network (H2020).
- Data@Exascale: Associate Team with the "Politehnica" University of Bucharest, Romania
- ANR OverFlow: data management for geo-distributed workflows on clouds
Research teams of the same theme :
- ALPINES - Algorithms and parallel tools for integrated numerical simulations
- AVALON - Algorithms and Software Architectures for Distributed and HPC Platforms
- DATAMOVE - Data Aware Large Scale Computing
- HIEPACS - High-End Parallel Algorithms for Challenging Numerical Simulations
- POLARIS - Performance analysis and Optimization of LARge Infrastructures and Systems
- ROMA - Optimisation des ressources : modèles, algorithmes et ordonnancement
- STORM - STatic Optimizations, Runtime Methods
- TADAAM - Topology-aware system-scale data management for high-performance computing