Real-time scheduling of multiple segment tasks

Kamhing Ho, James H. Rice, Jaideep Srivastava

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

The authors study the problem of on-line non-preemptive scheduling of multiple segment real-time tasks. Task segments alternate between using CPU and I/O resources. A task model is proposed which encompasses a wider class of tasks than models proposed earlier. Total slack available to a task is modeled as a new kind of virtual resource which is spent while waiting in queues for service by physical resources. Instead of developing new scheduling algorithms, the authors develop a class of slack distribution policies which use varying degrees of information about task structure and device utilization to budget task slack. Slack distribution policies are shown to improve the performance of all scheduling algorithms studied. Two key observations are: (1) slack distribution is helpful beyond a certain threshold of task arrival rate, and (2) algorithms which normally perform poorly are helped to a greater degree by slack distribution. A study of various scheduling algorithms for a constant value function reveals that all of them favor tasks with a large number of small segments to tasks with a small number of large segments. It is shown that the Moore ordering algorithm is not optimal for multiple segment tasks.

Original languageEnglish
Title of host publicationProceedings - IEEE Computer Society's International Computer Software & Applications Conference
PublisherPubl by IEEE
Pages680-686
Number of pages7
Publication statusPublished - 1990
Externally publishedYes
EventProceedings of the 14th Annual International Computer Software and Applications Conference - COMPSAC 90 - Chicago, IL, USA
Duration: 29 Oct 19902 Nov 1990

Other

OtherProceedings of the 14th Annual International Computer Software and Applications Conference - COMPSAC 90
CityChicago, IL, USA
Period29/10/902/11/90

Fingerprint

Scheduling algorithms
Scheduling
Program processors

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Ho, K., Rice, J. H., & Srivastava, J. (1990). Real-time scheduling of multiple segment tasks. In Proceedings - IEEE Computer Society's International Computer Software & Applications Conference (pp. 680-686). Publ by IEEE.

Real-time scheduling of multiple segment tasks. / Ho, Kamhing; Rice, James H.; Srivastava, Jaideep.

Proceedings - IEEE Computer Society's International Computer Software & Applications Conference. Publ by IEEE, 1990. p. 680-686.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Ho, K, Rice, JH & Srivastava, J 1990, Real-time scheduling of multiple segment tasks. in Proceedings - IEEE Computer Society's International Computer Software & Applications Conference. Publ by IEEE, pp. 680-686, Proceedings of the 14th Annual International Computer Software and Applications Conference - COMPSAC 90, Chicago, IL, USA, 29/10/90.
Ho K, Rice JH, Srivastava J. Real-time scheduling of multiple segment tasks. In Proceedings - IEEE Computer Society's International Computer Software & Applications Conference. Publ by IEEE. 1990. p. 680-686
Ho, Kamhing ; Rice, James H. ; Srivastava, Jaideep. / Real-time scheduling of multiple segment tasks. Proceedings - IEEE Computer Society's International Computer Software & Applications Conference. Publ by IEEE, 1990. pp. 680-686
@inproceedings{ad3bf9fb3abb4dda967e02a664296d3f,
title = "Real-time scheduling of multiple segment tasks",
abstract = "The authors study the problem of on-line non-preemptive scheduling of multiple segment real-time tasks. Task segments alternate between using CPU and I/O resources. A task model is proposed which encompasses a wider class of tasks than models proposed earlier. Total slack available to a task is modeled as a new kind of virtual resource which is spent while waiting in queues for service by physical resources. Instead of developing new scheduling algorithms, the authors develop a class of slack distribution policies which use varying degrees of information about task structure and device utilization to budget task slack. Slack distribution policies are shown to improve the performance of all scheduling algorithms studied. Two key observations are: (1) slack distribution is helpful beyond a certain threshold of task arrival rate, and (2) algorithms which normally perform poorly are helped to a greater degree by slack distribution. A study of various scheduling algorithms for a constant value function reveals that all of them favor tasks with a large number of small segments to tasks with a small number of large segments. It is shown that the Moore ordering algorithm is not optimal for multiple segment tasks.",
author = "Kamhing Ho and Rice, {James H.} and Jaideep Srivastava",
year = "1990",
language = "English",
pages = "680--686",
booktitle = "Proceedings - IEEE Computer Society's International Computer Software & Applications Conference",
publisher = "Publ by IEEE",

}

TY - GEN

T1 - Real-time scheduling of multiple segment tasks

AU - Ho, Kamhing

AU - Rice, James H.

AU - Srivastava, Jaideep

PY - 1990

Y1 - 1990

N2 - The authors study the problem of on-line non-preemptive scheduling of multiple segment real-time tasks. Task segments alternate between using CPU and I/O resources. A task model is proposed which encompasses a wider class of tasks than models proposed earlier. Total slack available to a task is modeled as a new kind of virtual resource which is spent while waiting in queues for service by physical resources. Instead of developing new scheduling algorithms, the authors develop a class of slack distribution policies which use varying degrees of information about task structure and device utilization to budget task slack. Slack distribution policies are shown to improve the performance of all scheduling algorithms studied. Two key observations are: (1) slack distribution is helpful beyond a certain threshold of task arrival rate, and (2) algorithms which normally perform poorly are helped to a greater degree by slack distribution. A study of various scheduling algorithms for a constant value function reveals that all of them favor tasks with a large number of small segments to tasks with a small number of large segments. It is shown that the Moore ordering algorithm is not optimal for multiple segment tasks.

AB - The authors study the problem of on-line non-preemptive scheduling of multiple segment real-time tasks. Task segments alternate between using CPU and I/O resources. A task model is proposed which encompasses a wider class of tasks than models proposed earlier. Total slack available to a task is modeled as a new kind of virtual resource which is spent while waiting in queues for service by physical resources. Instead of developing new scheduling algorithms, the authors develop a class of slack distribution policies which use varying degrees of information about task structure and device utilization to budget task slack. Slack distribution policies are shown to improve the performance of all scheduling algorithms studied. Two key observations are: (1) slack distribution is helpful beyond a certain threshold of task arrival rate, and (2) algorithms which normally perform poorly are helped to a greater degree by slack distribution. A study of various scheduling algorithms for a constant value function reveals that all of them favor tasks with a large number of small segments to tasks with a small number of large segments. It is shown that the Moore ordering algorithm is not optimal for multiple segment tasks.

UR - http://www.scopus.com/inward/record.url?scp=0025532935&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0025532935&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:0025532935

SP - 680

EP - 686

BT - Proceedings - IEEE Computer Society's International Computer Software & Applications Conference

PB - Publ by IEEE

ER -