Probabilistic backfilling

Avi Nissimov; Dror G. Feitelson

doi:10.1007/978-3-540-78699-3_6

Probabilistic backfilling

Avi Nissimov^*, Dror G. Feitelson

^*Corresponding author for this work

The Rachel and Selim Benin School of Engineering and Computer Science

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

11 Scopus citations

Abstract

Backfilling is a scheduling optimization that requires information about job runtimes to be known. Such information can come from either of two sources: estimates provided by users when the jobs are submitted, or predictions made by the system based on historical data regarding previous executions of jobs. In both cases, each job is assigned a precise prediction of how long it will run. We suggest that instead the whole distribution of the historical data be used. As a result, the whole backfilling framework shifts from a concrete plan for the future schedule to a probabilistic plan where jobs are backfilled based on the probability that they will terminate in time.

Original language	English
Title of host publication	Job Scheduling Strategies for Parallel Processing - 13th International Workshop, JSSPP 2007, Revised Papers
Pages	102-115
Number of pages	14
DOIs	https://doi.org/10.1007/978-3-540-78699-3_6
State	Published - 2008
Event	13th International Workshop on Job Scheduling Strategies for Parallel Processing, JSSPP 2007 - Seattle, WA, United States Duration: 17 Jun 2006 → 17 Jun 2006

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	4942 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	13th International Workshop on Job Scheduling Strategies for Parallel Processing, JSSPP 2007
Country/Territory	United States
City	Seattle, WA
Period	17/06/06 → 17/06/06

Access to Document

10.1007/978-3-540-78699-3_6

Cite this

Nissimov, A., & Feitelson, D. G. (2008). Probabilistic backfilling. In Job Scheduling Strategies for Parallel Processing - 13th International Workshop, JSSPP 2007, Revised Papers (pp. 102-115). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4942 LNCS). https://doi.org/10.1007/978-3-540-78699-3_6

@inproceedings{c6fe37711c5848a99851b16fc6da1237,

title = "Probabilistic backfilling",

abstract = "Backfilling is a scheduling optimization that requires information about job runtimes to be known. Such information can come from either of two sources: estimates provided by users when the jobs are submitted, or predictions made by the system based on historical data regarding previous executions of jobs. In both cases, each job is assigned a precise prediction of how long it will run. We suggest that instead the whole distribution of the historical data be used. As a result, the whole backfilling framework shifts from a concrete plan for the future schedule to a probabilistic plan where jobs are backfilled based on the probability that they will terminate in time.",

author = "Avi Nissimov and Feitelson, {Dror G.}",

year = "2008",

doi = "10.1007/978-3-540-78699-3_6",

language = "אנגלית",

isbn = "3540786988",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

pages = "102--115",

booktitle = "Job Scheduling Strategies for Parallel Processing - 13th International Workshop, JSSPP 2007, Revised Papers",

note = "13th International Workshop on Job Scheduling Strategies for Parallel Processing, JSSPP 2007 ; Conference date: 17-06-2006 Through 17-06-2006",

}

Nissimov, A & Feitelson, DG 2008, Probabilistic backfilling. in Job Scheduling Strategies for Parallel Processing - 13th International Workshop, JSSPP 2007, Revised Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 4942 LNCS, pp. 102-115, 13th International Workshop on Job Scheduling Strategies for Parallel Processing, JSSPP 2007, Seattle, WA, United States, 17/06/06. https://doi.org/10.1007/978-3-540-78699-3_6

Probabilistic backfilling. / Nissimov, Avi; Feitelson, Dror G.
Job Scheduling Strategies for Parallel Processing - 13th International Workshop, JSSPP 2007, Revised Papers. 2008. p. 102-115 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4942 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Probabilistic backfilling

AU - Nissimov, Avi

AU - Feitelson, Dror G.

PY - 2008

Y1 - 2008

N2 - Backfilling is a scheduling optimization that requires information about job runtimes to be known. Such information can come from either of two sources: estimates provided by users when the jobs are submitted, or predictions made by the system based on historical data regarding previous executions of jobs. In both cases, each job is assigned a precise prediction of how long it will run. We suggest that instead the whole distribution of the historical data be used. As a result, the whole backfilling framework shifts from a concrete plan for the future schedule to a probabilistic plan where jobs are backfilled based on the probability that they will terminate in time.

AB - Backfilling is a scheduling optimization that requires information about job runtimes to be known. Such information can come from either of two sources: estimates provided by users when the jobs are submitted, or predictions made by the system based on historical data regarding previous executions of jobs. In both cases, each job is assigned a precise prediction of how long it will run. We suggest that instead the whole distribution of the historical data be used. As a result, the whole backfilling framework shifts from a concrete plan for the future schedule to a probabilistic plan where jobs are backfilled based on the probability that they will terminate in time.

UR - http://www.scopus.com/inward/record.url?scp=43149103700&partnerID=8YFLogxK

U2 - 10.1007/978-3-540-78699-3_6

DO - 10.1007/978-3-540-78699-3_6

M3 - ???researchoutput.researchoutputtypes.contributiontobookanthology.conference???

AN - SCOPUS:43149103700

SN - 3540786988

SN - 9783540786986

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 102

EP - 115

BT - Job Scheduling Strategies for Parallel Processing - 13th International Workshop, JSSPP 2007, Revised Papers

T2 - 13th International Workshop on Job Scheduling Strategies for Parallel Processing, JSSPP 2007

Y2 - 17 June 2006 through 17 June 2006

ER -

Probabilistic backfilling

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this