posted on 2019-03-26, 13:54authored byTim Suss, Nils Doring, Ramy Gad, Lars NagelLars Nagel, Andre Brinkmann, Dustin Feld, Eric Schricker, Thomas Soddemann
In recent years, the number of processing units per compute node has
been increasing. In order to utilize all or most of the available resources of a highperformance computing cluster, at least some of its nodes will have to be shared by
several applications at the same time. Yet, even if jobs are co-scheduled on a node,
it can happen that high performance resources remain idle, although there are jobs
that could make use of them (e. g., if the resource was temporarily blocked when
the job was started). Heterogeneous schedulers, which schedule tasks for different
devices, can bind jobs to resources in a way that can significantly reduce the idle
time. Typically, such schedulers make their decisions based on a static strategy.
We investigate the impact of allowing a heterogeneous scheduler to modify its
strategy at runtime. For a set of applications, we determine the makespan and show
how it is influenced by four different scheduling strategies. A strategy tailored to
one use case can be disastrous in another one and can consequently even result in a
slowdown - in our experiments of up to factor 2.5.
History
School
Science
Department
Computer Science
Published in
1st Workshop on Co-Scheduling of HPC Applications, COSH@HiPEAC 2016 (extended versions)
Advances in Parallel Computing: Co-Scheduling of HPC Applications
Volume
28
Citation
SUSS, T. ... et al., 2017. Impact of the scheduling strategy in heterogeneous systems that provide co-scheduling. IN: Trinitis, C. and Weidendorfer , J. (eds.) Co-Scheduling of HPC Applications. Amsterdam: IOS Press, pp. 142-162.
This work is made available according to the conditions of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) licence. Full details of this licence are available at: https://creativecommons.org/licenses/by-nc-nd/4.0/
Acceptance date
2016-11-16
Publication date
2017
Notes
This paper was initially presented at the 1st Workshop on Co-Scheduling of HPC Applications, COSH@HiPEAC 2016, Prague, Czech Republic, Jan 19th 2016. This paper is published with kind permission of IOS Press. The publication is available at IOS Press through http://doi.org/10.3233/978-1-61499-730-6-142