Algorithms for hierarchical and semi-partitioned parallel scheduling - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Journal Articles Journal of Computer and System Sciences Year : 2021

Algorithms for hierarchical and semi-partitioned parallel scheduling

Abstract

We propose a model for scheduling jobs in a parallel machine setting that takes into account the cost of migrations by assuming that the processing time of a job may depend on the specific set of machines among which the job is migrated. For the makespan minimization objective, the model generalizes classical scheduling problems such as unrelated parallel machine scheduling, as well as novel ones such as semi-partitioned and clustered scheduling. In the case of a hierarchical family of machines, we derive a compact integer linear programming formulation of the problem and leverage its fractional relaxation to obtain a polynomial-time 2- approximation algorithm. Extensions that incorporate memory capacity constraints are also discussed.
Fichier principal
Vignette du fichier
affinity-ipdps.pdf (326.37 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-03498319 , version 1 (21-12-2021)

Identifiers

Cite

Vincenzo Bonifaci, Gianlorenzo d'Angelo, Alberto Marchetti-Spaccamela. Algorithms for hierarchical and semi-partitioned parallel scheduling. Journal of Computer and System Sciences, 2021, 120, pp.116-136. ⟨10.1016/j.jcss.2021.03.006⟩. ⟨hal-03498319⟩

Collections

INRIA INRIA2
9 View
46 Download

Altmetric

Share

Gmail Facebook X LinkedIn More