One of the main activities of the SPEEDUP Society is the organization of an annual workshop on high performance computing and related topics for the Swiss scientific community. Each year the SPEEDUP workshop addresses the latest advances in high computing technology, focussing on a specific application field, such as molecular dynamics, environmental impacts, bioinformatics, or [...]
Whamcloud and CSCS Sign Lustre Support Agreement
Whamcloud and CSCS have announced a Lustre support agreement. CSCS has a Lustre-based scratch file system for their flagship supercomputer system, Monte Rosa, which is a 47,872 core Cray XE6. The file system uses the latest Intel Sandy Bridge server hardware, FDR InfiniBand and Lustre v2.2. “CSCS is one of the shining stars in the [...]
Slidecast: Running GPFS over Obsidian LongBow, Hussein N. Harake
Hussein N. Harake presented at the HPC Advisory Council in Lugano how CSCS is moving without service disruption data from the old location in Manno to a new building in Lugano. Both locations are a few kilometers kilometers away from each other. To to this CSCS is running GPFS over Obsidian LongBow. Introduction: Move from [...]
Slidecast: New HSM over GPFS/TSM at CSCS
Stefano Gorini (CSCS) presented at the HPC Advisory Council in Lugano the storage of CSCS based on a GPFS / HSM solution. The GPFS filesystem at CSCS QFS/SAM-FS to GPFS TSM Migration Data topology (file size range as a function of last access time) TSM/HSM Open issues Future plans (please apologize for the low audio [...]
Slidecast: HPC Disaster Recovery at CSCS, Hussein N. Harake
At the last HPC Advisory Council Workshop in Lugano Hussein N. Harake presented how the storage disaster recovery is implemented at CSCS based on HSM solution of IBM. Contents: Overview of CSCS resources Disaster Recovery Requirements Data & File Systems Archiving and Backup Solution Backup & Restore The challenging part





