Publication in Conference Proceedings/Workshop.
Complex applications and workflows needs are often exclusively expressed in terms of computational resources on HPC systems. In many cases, other resources like storage or network are not allocatable and are shared across the entire HPC system. By looking at the storage resources in particular, any workflow or application should be able to select both its preferred data manager and its required storage capability or capacity. To achieve such a goal, new mechanisms should be introduced. In this work, we present such a tool that dynamically provisions a data management system on top of storage devices. We propose a proof-of-concept that is able to deploy, on-demand, a parallel file-system across intermediate storage nodes on a Cray XC50 system. We show how this mechanism can be easily extended to support more data managers and any type of intermediate storage. Finally, we evaluate the performance of the provisioned storage system with a set of benchmarks.
Details
Publication Type: Workshop
Year of Publication: 2020
Authors: François Tessier, Maxime Martinasso, Matteo Chesi, Mark Klein, Miguel Gila (CSCS, ETH Zürich)
Publisher: 2020 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW) – Workshop on High-Performance Storage
Place published: New Orleans, LA, USA, USA
Keywords: Dynamic scheduling, Buffer storage,Tools,Containers,
Hardware,Task analysis, Computer architecture.
Publication link: https://ieeexplore.ieee.org/document/9150373
DOI: 10.1109/IPDPSW50202.2020.00173
Slides of the workshop: http://snir.cs.illinois.edu/HPS/slides/HPS_Tessier.pdf
Leave a Reply