heSRPT: Parallel scheduling to minimize mean slowdown
Performance Evaluation · 2020-12 · 144 · 102147
DOI: 10.1016/j.peva.2020.102147 ↗Kurz gefasst
Fachartikel in der Zeitschrift Performance Evaluation, 2020-12. Verfasst von Benjamin Berg und 2 weiteren Autor:innen. Die Arbeit wurde bisher 8-mal zitiert (Crossref) und nennt 45 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0). Gefördert von: Facebook, National Science Foundation, Google.
Abstract (Originalsprache)
Modern data centers serve workloads which can exploit parallelism. When a job parallelizes across multiple servers it completes more quickly. However, it is unclear how to share a limited number of servers between many parallelizable jobs. In this paper we consider a typical scenario where a data center composed of N servers will be tasked with completing a set of M parallelizable jobs. Typically, M is much smaller than N. In our scenario, each job consists of some amount of inherent work which we refer to as a job's size. We assume that job sizes are known up front to the system, and each job can utilize any number of servers at any moment in time. These assumptions are reasonable for many parallelizable workloads such as training neural networks using TensorFlow [2]. Our goal in this paper is to allocate servers to jobs so as to minimize the mean slowdown across all jobs, where the slowdown of a job is the job's completion time divided by its running time if given exclusive access to all N servers. Slowdown measures how a job was interfered with by other jobs in the system, and is often the metric of interest in the theoretical parallel scheduling literature (where it is also called stretch), as well as the HPC community (where it is called expansion factor).
Förderung
- National Science Foundation
Zitieren
APA
Berg, B., Vesilo, R., & Harchol-Balter, M. (2020). heSRPT: Parallel scheduling to minimize mean slowdown. Performance Evaluation, 144, 102147. https://doi.org/10.1016/j.peva.2020.102147
Vancouver
Berg B, Vesilo R, Harchol-Balter M. heSRPT: Parallel scheduling to minimize mean slowdown. Performance Evaluation. 2020;144:102147. doi:10.1016/j.peva.2020.102147
BibTeX
@article{Berg2020,
author = {Berg, Benjamin and Vesilo, Rein and Harchol-Balter, Mor},
title = {heSRPT: Parallel scheduling to minimize mean slowdown},
journal = {Performance Evaluation},
year = {2020},
volume = {144},
pages = {102147},
doi = {10.1016/j.peva.2020.102147}
}
EUVIMED