A5055399769- Advanced Data Storage Technologies
- Distributed and Parallel Computing Systems
- Parallel Computing and Optimization Techniques
- Caching and Content Delivery
- Cloud Computing and Resource Management
- Scientific Computing and Data Management
- Distributed systems and fault tolerance
- Advanced Database Systems and Queries
- Data Management and Algorithms
Commissariat à l'Énergie Atomique et aux Énergies Alternatives
2021-2024
Laboratoire de l'Informatique du Parallélisme
2015-2023
CEA DAM Île-de-France
2021-2023
CEA Paris-Saclay
2021
Université de Versailles Saint-Quentin-en-Yvelines
2018
Laboratoire d’informatique Parallélisme Réseaux Algorithmes Distribués
2015-2017
Institut Lavoisier de Versailles
2015
The emergence of Exascale machines in HPC will have the foreseen consequence putting more pressure on storage systems place, not only terms capacity but also bandwidth and latency. With limited budget we cannot imagine using class memory, which leads to use a heterogeneous tiered hierarchy. In order make most efficient high performance tier this hierarchy, need be able place user data right at time. paper, assume 2-tier hierarchy with archival tier. Files are placed creation time moved once...
Using simulation to study the behavior of large-scale data storage systems becomes capital predict performance and reliability any them at a lower cost. This helps take right decisions before system development deployment. OGSSim is tool for large heterogeneous that uses parallelism provide informations about such in reduced time. It ZeroMQ communication library implement not only but also synchronization functions between generated threads. These points occur during requests parallel...
In this paper, an open and generic storage simulator is proposed. It simulates with accuracy multi-tiered systems based on heterogeneous devices including HDDs, SSDs the connecting buses. The target simulated sys- tem constructed from hardware configuration input, then sent t
Actual storage systems are very large, with complex and distributed architectural configurations, composed of various technologies devices. However, simulation, analysis evaluation tools in the literature do not handle this design these heterogeneous components. This paper presents OGSSim (Open Generic Storage Simulation tool): a new simulation tool for such systems. Being generic to all devices open diverse management strategies architecture layouts, it fulfills needs term...
Modern disks are very large (SSDs, HDDs) and their capacities will certainly increase in the future. Storage systems use an important number of such devices to compose storage pools fulfil capacity demands. The result is a higher probability failure longer reconstruction duration. Consequently, whole system penalized as response time second generate data loss. In this paper, we propose new method based on block shifting layout which increases efficiency RAID declustered improves its...
The emergence of Exascale machines in HPC will have the foreseen consequence putting more pressure on storage systems place, not only terms capacity but also bandwidth and latency. With limited budget we cannot imagine using class memory, which leads to use a heterogeneous tiered hierarchy. In order make most efficient high performance tier this hierarchy, need be able place user data right at time. paper, assume 2-tier hierarchy with archival tier. Files are placed creation time moved once...
Storage systems capacity provided by data centers do not cease to increase, currently reaching the exabyte scale using thousands of disks. In this way, question resiliency such becomes critical, avoid loss and reduce impact reconstruction process on access time. We propose SD2S, a method create placement scheme for declustered RAID organizations, based shifting placement. It consists in calculation degree matrices, which represent distance between source sets each couple physical disks, thus...
Simulation is the most appropriate technique to evaluate performance of current data storage systems and predict it for future ones as part centers or cloud infrastructures. It assesses potential a system meet users requirements in terms capacity, devices heterogeneity, delivered robustness. We developed simulation tool called OGSSim address efficiently these criteria within reduced execution time. But number threads on test machine put an upper bound size simulated systems. To push this...
Simulation is the most appropriate technique to evaluate performance of current data storage systems and predict it for future ones as part centers or cloud infrastructures. It assesses potential a system meet users requirements in terms capacity, devices heterogeneity, delivered robustness. We developed simulation tool called OGSSim address efficiently these criteria within reduced execution time. But number threads on test machine put an upper bound size simulated systems. To push this...