Job characteristics of a production parallel scientific workload on the NASA Ames iPSC/860

Dror G. Feitelson; Bill Nitzberg

doi:10.1007/3-540-60153-8_38

Job characteristics of a production parallel scientific workload on the NASA Ames iPSC/860

Dror G. Feitelson, Bill Nitzberg

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

139 Scopus citations

Abstract

Statistics of a parallel workload on a 128-node iPSC/860 located at NASA Ames are presented. It is shown that while the number of sequential jobs dominates the number of parallel jobs, most of the resources (measured in node-seconds) were consumed by parallel jobs. Moreover, most of the sequential jobs were for system administration. The average runtime of jobs grew with the number of nodes used, so the total resource requirements of large parallel jobs were larger by more than the number of nodes they used. The job submission rate during peak day activity was somewhat lower than one every two minutes, and the average job size was small. At night, submission rate was low but job sizes and system utilization were high, mainly due to NQS. Submission rate and utilization over the weekend were lower than on weekdays. The overall utilization was 50%, after accounting for downtime. About 2/3 of the applications were executed repeatedly, some for a significant number of times.

Original language	English
Title of host publication	Job Scheduling Strategies for Parallel Processing - IPPS 1995 Workshop, Proceedings
Editors	Dror G. Feitelson, Larry Rudolph
Publisher	Springer Verlag
Pages	337-360
Number of pages	24
ISBN (Print)	3540601538, 9783540601531
DOIs	https://doi.org/10.1007/3-540-60153-8_38
State	Published - 1995
Externally published	Yes
Event	Workshop on Job Scheduling Strategies for Parallel Processing held as a prelude to the IPPS 1995 Conference - Santa Barbara, United States Duration: 25 Apr 1995 → 25 Apr 1995

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	949
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	Workshop on Job Scheduling Strategies for Parallel Processing held as a prelude to the IPPS 1995 Conference
Country/Territory	United States
City	Santa Barbara
Period	25/04/95 → 25/04/95

Bibliographical note

Publisher Copyright:
© 1995, Springer Verlag. All rights reserved.

Access to Document

10.1007/3-540-60153-8_38

Cite this

Feitelson, D. G., & Nitzberg, B. (1995). Job characteristics of a production parallel scientific workload on the NASA Ames iPSC/860. In D. G. Feitelson, & L. Rudolph (Eds.), Job Scheduling Strategies for Parallel Processing - IPPS 1995 Workshop, Proceedings (pp. 337-360). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 949). Springer Verlag. https://doi.org/10.1007/3-540-60153-8_38

Feitelson, Dror G. ; Nitzberg, Bill. / Job characteristics of a production parallel scientific workload on the NASA Ames iPSC/860. Job Scheduling Strategies for Parallel Processing - IPPS 1995 Workshop, Proceedings. editor / Dror G. Feitelson ; Larry Rudolph. Springer Verlag, 1995. pp. 337-360 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{183b3996026940bc95a1fb6ded99162b,

title = "Job characteristics of a production parallel scientific workload on the NASA Ames iPSC/860",

abstract = "Statistics of a parallel workload on a 128-node iPSC/860 located at NASA Ames are presented. It is shown that while the number of sequential jobs dominates the number of parallel jobs, most of the resources (measured in node-seconds) were consumed by parallel jobs. Moreover, most of the sequential jobs were for system administration. The average runtime of jobs grew with the number of nodes used, so the total resource requirements of large parallel jobs were larger by more than the number of nodes they used. The job submission rate during peak day activity was somewhat lower than one every two minutes, and the average job size was small. At night, submission rate was low but job sizes and system utilization were high, mainly due to NQS. Submission rate and utilization over the weekend were lower than on weekdays. The overall utilization was 50%, after accounting for downtime. About 2/3 of the applications were executed repeatedly, some for a significant number of times.",

author = "Feitelson, {Dror G.} and Bill Nitzberg",

note = "Publisher Copyright: {\textcopyright} 1995, Springer Verlag. All rights reserved.; Workshop on Job Scheduling Strategies for Parallel Processing held as a prelude to the IPPS 1995 Conference ; Conference date: 25-04-1995 Through 25-04-1995",

year = "1995",

doi = "10.1007/3-540-60153-8_38",

language = "אנגלית",

isbn = "3540601538",

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

publisher = "Springer Verlag",

pages = "337--360",

editor = "Feitelson, {Dror G.} and Larry Rudolph",

booktitle = "Job Scheduling Strategies for Parallel Processing - IPPS 1995 Workshop, Proceedings",

address = "גרמניה",

}

Feitelson, DG & Nitzberg, B 1995, Job characteristics of a production parallel scientific workload on the NASA Ames iPSC/860. in DG Feitelson & L Rudolph (eds), Job Scheduling Strategies for Parallel Processing - IPPS 1995 Workshop, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 949, Springer Verlag, pp. 337-360, Workshop on Job Scheduling Strategies for Parallel Processing held as a prelude to the IPPS 1995 Conference, Santa Barbara, United States, 25/04/95. https://doi.org/10.1007/3-540-60153-8_38

Job characteristics of a production parallel scientific workload on the NASA Ames iPSC/860. / Feitelson, Dror G.; Nitzberg, Bill.
Job Scheduling Strategies for Parallel Processing - IPPS 1995 Workshop, Proceedings. ed. / Dror G. Feitelson; Larry Rudolph. Springer Verlag, 1995. p. 337-360 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 949).

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

TY - GEN

T1 - Job characteristics of a production parallel scientific workload on the NASA Ames iPSC/860

AU - Feitelson, Dror G.

AU - Nitzberg, Bill

PY - 1995

Y1 - 1995

N2 - Statistics of a parallel workload on a 128-node iPSC/860 located at NASA Ames are presented. It is shown that while the number of sequential jobs dominates the number of parallel jobs, most of the resources (measured in node-seconds) were consumed by parallel jobs. Moreover, most of the sequential jobs were for system administration. The average runtime of jobs grew with the number of nodes used, so the total resource requirements of large parallel jobs were larger by more than the number of nodes they used. The job submission rate during peak day activity was somewhat lower than one every two minutes, and the average job size was small. At night, submission rate was low but job sizes and system utilization were high, mainly due to NQS. Submission rate and utilization over the weekend were lower than on weekdays. The overall utilization was 50%, after accounting for downtime. About 2/3 of the applications were executed repeatedly, some for a significant number of times.

AB - Statistics of a parallel workload on a 128-node iPSC/860 located at NASA Ames are presented. It is shown that while the number of sequential jobs dominates the number of parallel jobs, most of the resources (measured in node-seconds) were consumed by parallel jobs. Moreover, most of the sequential jobs were for system administration. The average runtime of jobs grew with the number of nodes used, so the total resource requirements of large parallel jobs were larger by more than the number of nodes they used. The job submission rate during peak day activity was somewhat lower than one every two minutes, and the average job size was small. At night, submission rate was low but job sizes and system utilization were high, mainly due to NQS. Submission rate and utilization over the weekend were lower than on weekdays. The overall utilization was 50%, after accounting for downtime. About 2/3 of the applications were executed repeatedly, some for a significant number of times.

UR - http://www.scopus.com/inward/record.url?scp=84947791890&partnerID=8YFLogxK

U2 - 10.1007/3-540-60153-8_38

DO - 10.1007/3-540-60153-8_38

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

AN - SCOPUS:84947791890

SN - 3540601538

SN - 9783540601531

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

SP - 337

EP - 360

BT - Job Scheduling Strategies for Parallel Processing - IPPS 1995 Workshop, Proceedings

A2 - Feitelson, Dror G.

A2 - Rudolph, Larry

PB - Springer Verlag

T2 - Workshop on Job Scheduling Strategies for Parallel Processing held as a prelude to the IPPS 1995 Conference

Y2 - 25 April 1995 through 25 April 1995

ER -

Feitelson DG, Nitzberg B. Job characteristics of a production parallel scientific workload on the NASA Ames iPSC/860. In Feitelson DG, Rudolph L, editors, Job Scheduling Strategies for Parallel Processing - IPPS 1995 Workshop, Proceedings. Springer Verlag. 1995. p. 337-360. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/3-540-60153-8_38

Job characteristics of a production parallel scientific workload on the NASA Ames iPSC/860

Abstract

Publication series

Conference

Bibliographical note

Access to Document

Other files and links

Fingerprint

Cite this