Dynamic Fractional Resource Scheduling for HPC Workloads

Abstract : We propose a novel job scheduling approach for homogeneous cluster computing platforms. Its key feature is the use of virtual machine technology for sharing resources in a precise and controlled manner. We justify our approach and propose several job scheduling algorithms. We present results obtained in simulations for synthetic and real-world High Performance Computing (HPC) workloads, in which we compare our proposed algorithms with standard batch scheduling algorithms. We find that our approach widely outperforms batch scheduling. We also identify a few promising algorithms that perform well across most experimental scenarios. Our results demonstrate that virtualization technology coupled with lightweight scheduling strategies affords dramatic improvements in performance for HPC workloads.
Complete list of metadatas

Cited literature [48 references]  Display  Hide  Download

https://hal.inria.fr/inria-00527068
Contributor : Frédéric Vivien <>
Submitted on : Monday, October 18, 2010 - 9:02:57 AM
Last modification on : Saturday, April 21, 2018 - 1:27:08 AM
Long-term archiving on : Wednesday, January 19, 2011 - 2:37:52 AM

File

IPDPS2010-SVC.pdf
Publisher files allowed on an open archive

Identifiers

Collections

Citation

Mark Stillwell, Frédéric Vivien, Henri Casanova. Dynamic Fractional Resource Scheduling for HPC Workloads. IEEE International Symposium on Parallel & Distributed Processing (IPDPS), 2010, Apr 2010, Atlanta, GA, United States. pp.1-12, ⟨10.1109/IPDPS.2010.5470356⟩. ⟨inria-00527068⟩

Share

Metrics

Record views

450

Files downloads

229