User-Aware Metrics for Measuring Quality of Parallel Job
Schedules

D 2015

User-Aware Metrics for Measuring Quality of Parallel Job Schedules

TÓTH, Šimon and Dalibor KLUSÁČEK

Basic information

Original name

User-Aware Metrics for Measuring Quality of Parallel Job Schedules

Authors

TÓTH, Šimon (203 Czech Republic, guarantor, belonging to the institution) and Dalibor KLUSÁČEK (203 Czech Republic, belonging to the institution)

Edition

1. vyd. Switzerland, Job Scheduling Strategies for Parallel Processing, p. 90-107, 18 pp. 2015

Publisher

Springer, Lecture Notes in Computer Science 8828

Other information

Language

English

Type of outcome

Stať ve sborníku

Field of Study

10201 Computer sciences, information science, bioinformatics

Country of publisher

Germany

Confidentiality degree

není předmětem státního či obchodního tajemství

Publication form

printed version "print"

References:

URL

Impact factor

Impact factor: 0.402 in 2005

RIV identification code

RIV/00216224:14330/15:00080586

Organization unit

Faculty of Informatics

ISBN

978-3-319-15788-7

ISSN

DOI

http://dx.doi.org/10.1007/978-3-319-15789-4_6

UT WoS

000355729800006

Keywords in English

Grid; Performance evaluation; Metrics; Queue-based scheduling; Fairness; User-aware scheduling

Abstract

V originále

The work presented in this paper is motivated by the challenges in the design of scheduling algorithms for the Czech National Grid MetaCentrum. One of the most notable problems is our inability to efficiently analyze the quality of schedules. While it is still possible to observe and measure certain aspects of generated schedules using various metrics, it is very challenging to choose a set of metrics that would be representative when measuring the schedule quality. Without quality quantification (either relative, or absolute), we have no way to determine the impact of new algorithms and configurations on the schedule quality, prior to their deployment in a production service. The only two options we are left with is to either use expert assessment or to simply deploy new solutions into production and observe their impact on user satisfaction. To approach this problem, we have designed a novel user-aware model and a~metric that can overcome the presented issues by evaluating the quality on a~user level. The model assigns an expected end time (EET) to each job based on a fair partitioning of the system resources, modeling users expectations. Using this calculated EET we can then compare generated schedules in detail, while also being able to adequately visualize schedule artifacts, allowing an expert to further analyze them. Moreover, we present how coupling this model with a job scheduling simulator gives us the ability to do an in-depth evaluation of scheduling algorithms before they are deployed into a production environment.

Links

GAP202/12/0306, research and development project

Name: Dyschnet - Dynamické plánování a rozvrhování výpočetních a síťových zdrojů (Acronym: Dyschnet)

Investor: Czech Science Foundation

MUNI/A/1159/2014, interní kód MU

Name: Rozsáhlé výpočetní systémy: modely, aplikace a verifikace IV.

Investor: Masaryk University, Category A

Citovat

TÓTH, Šimon and Dalibor KLUSÁČEK. User-Aware Metrics for Measuring Quality of Parallel Job Schedules. In Walfredo Cirne and Narayan Desai. Job Scheduling Strategies for Parallel Processing. 1st ed. Switzerland: Springer, Lecture Notes in Computer Science 8828, 2015, p. 90-107. ISBN 978-3-319-15788-7. Available from: https://dx.doi.org/10.1007/978-3-319-15789-4_6.

@inproceedings{1186294,
   author = {Tóth, Šimon and Klusáček, Dalibor},
   address = {Switzerland},
   booktitle = {Job Scheduling Strategies for Parallel Processing},
   doi = {http://dx.doi.org/10.1007/978-3-319-15789-4_6},
   edition = {1},
   editor = {Walfredo Cirne and Narayan Desai},
   keywords = {Grid; Performance evaluation; Metrics; Queue-based scheduling; Fairness; User-aware scheduling},
   howpublished = {tištěná verze "print"},
   language = {eng},
   location = {Switzerland},
   isbn = {978-3-319-15788-7},
   pages = {90-107},
   publisher = {Springer, Lecture Notes in Computer Science 8828},
   title = {User-Aware Metrics for Measuring Quality of Parallel Job Schedules},
   url = {http://www.cs.huji.ac.il/~feit/parsched/jsspp14/p3-toth.pdf},
   year = {2015}
}

TY  - JOUR
ID  - 1186294
AU  - Tóth, Šimon - Klusáček, Dalibor
PY  - 2015
TI  - User-Aware Metrics for Measuring Quality of Parallel Job Schedules
PB  - Springer, Lecture Notes in Computer Science 8828
CY  - Switzerland
SN  - 9783319157887
KW  - Grid
KW  - Performance evaluation
KW  - Metrics
KW  - Queue-based scheduling
KW  - Fairness
KW  - User-aware scheduling
UR  - http://www.cs.huji.ac.il/~feit/parsched/jsspp14/p3-toth.pdf
L2  - http://www.cs.huji.ac.il/~feit/parsched/jsspp14/p3-toth.pdf
N2  - The work presented in this paper is motivated by the challenges in the design of scheduling algorithms for the Czech National Grid MetaCentrum. One of the most notable problems is our inability to efficiently analyze the quality of schedules. While it is still possible to observe and measure certain aspects of generated schedules using various metrics, it is very challenging to choose a set of metrics that would be representative when measuring the schedule quality. Without quality quantification (either relative, or absolute), we have no way to determine the impact of new algorithms and configurations on the schedule quality, prior to their deployment in a production service. The only two options we are left with is to either use expert assessment or to simply deploy new solutions into production and observe their impact on user satisfaction. To approach this problem, we have designed a novel user-aware model and a~metric that can overcome the presented issues by evaluating the quality on a~user level. The model assigns an expected end time (EET) to each job based on a fair partitioning of the system resources, modeling users expectations. Using this calculated EET we can then compare generated schedules in detail, while also being able to adequately visualize schedule artifacts, allowing an expert to further analyze them. Moreover, we present how coupling this model with a job scheduling simulator gives us the ability to do an in-depth evaluation of scheduling algorithms before they are deployed into a production environment.
ER  -

TÓTH, Šimon and Dalibor KLUSÁČEK. User-Aware Metrics for Measuring Quality of Parallel Job Schedules. In Walfredo Cirne and Narayan Desai. \textit{Job Scheduling Strategies for Parallel Processing}. 1st ed. Switzerland: Springer, Lecture Notes in Computer Science 8828, 2015, p.~90-107. ISBN~978-3-319-15788-7. Available from: https://dx.doi.org/10.1007/978-3-319-15789-4\_{}6.

Detailed Information on Publication Record