On the competitiveness of scheduling dynamically injected tasks on processes prone to crashes and restarts

Chryssis Georgiou, Dariusz R. Kowalski

Research output: Contribution to journalArticle

Abstract

To identify the tradeoffs between efficiency and fault-tolerance in dynamic cooperative computing, we initiate the study of a task performing problem under dynamic processes' crashes/restarts and task injections. The system consists of n message-passing processes which, subject to dynamic crashes and restarts, cooperate in performing tasks that are continuously and dynamically injected to the system. Tasks are not known a priori to the processes. This problem abstracts todays Internet-based computations, such as Grid computing and cloud services, where tasks are generated dynamically and different tasks may become known to different processes. We measure performance in terms of the number of pending tasks, and as such it can be directly compared with the optimum number obtained under the same crash-restart-injection pattern by the best off-line algorithm. Hence, we view the problem as an online problem and we pursue competitive analysis. We propose several deterministic algorithmic solutions to the considered problem under different information models and correctness criteria, and we argue that their performance is close to the best possible offline solutions. We also prove negative results that open interesting research directions.

Original languageEnglish (US)
Pages (from-to)94-107
Number of pages14
JournalJournal of Parallel and Distributed Computing
Volume84
DOIs
StatePublished - Aug 13 2015
Externally publishedYes

Fingerprint

Restart
Competitiveness
Crash
Scheduling
Grid computing
Message passing
Fault tolerance
Injection
Competitive Analysis
Internet
Dynamic Process
Grid Computing
Message Passing
Fault Tolerance
Performance Measures
Correctness
Trade-offs
Computing
Line

Keywords

  • Competitive analysis
  • Distributed algorithms
  • Dynamic task injection
  • Processor crashes and restarts
  • Task execution

ASJC Scopus subject areas

  • Software
  • Theoretical Computer Science
  • Hardware and Architecture
  • Computer Networks and Communications
  • Artificial Intelligence

Cite this

On the competitiveness of scheduling dynamically injected tasks on processes prone to crashes and restarts. / Georgiou, Chryssis; Kowalski, Dariusz R.

In: Journal of Parallel and Distributed Computing, Vol. 84, 13.08.2015, p. 94-107.

Research output: Contribution to journalArticle

@article{e1882410b87b444db11a0c2c2e7bb424,
title = "On the competitiveness of scheduling dynamically injected tasks on processes prone to crashes and restarts",
abstract = "To identify the tradeoffs between efficiency and fault-tolerance in dynamic cooperative computing, we initiate the study of a task performing problem under dynamic processes' crashes/restarts and task injections. The system consists of n message-passing processes which, subject to dynamic crashes and restarts, cooperate in performing tasks that are continuously and dynamically injected to the system. Tasks are not known a priori to the processes. This problem abstracts todays Internet-based computations, such as Grid computing and cloud services, where tasks are generated dynamically and different tasks may become known to different processes. We measure performance in terms of the number of pending tasks, and as such it can be directly compared with the optimum number obtained under the same crash-restart-injection pattern by the best off-line algorithm. Hence, we view the problem as an online problem and we pursue competitive analysis. We propose several deterministic algorithmic solutions to the considered problem under different information models and correctness criteria, and we argue that their performance is close to the best possible offline solutions. We also prove negative results that open interesting research directions.",
keywords = "Competitive analysis, Distributed algorithms, Dynamic task injection, Processor crashes and restarts, Task execution",
author = "Chryssis Georgiou and Kowalski, {Dariusz R.}",
year = "2015",
month = "8",
day = "13",
doi = "10.1016/j.jpdc.2015.07.007",
language = "English (US)",
volume = "84",
pages = "94--107",
journal = "Journal of Parallel and Distributed Computing",
issn = "0743-7315",
publisher = "Academic Press Inc.",

}

TY - JOUR

T1 - On the competitiveness of scheduling dynamically injected tasks on processes prone to crashes and restarts

AU - Georgiou, Chryssis

AU - Kowalski, Dariusz R.

PY - 2015/8/13

Y1 - 2015/8/13

N2 - To identify the tradeoffs between efficiency and fault-tolerance in dynamic cooperative computing, we initiate the study of a task performing problem under dynamic processes' crashes/restarts and task injections. The system consists of n message-passing processes which, subject to dynamic crashes and restarts, cooperate in performing tasks that are continuously and dynamically injected to the system. Tasks are not known a priori to the processes. This problem abstracts todays Internet-based computations, such as Grid computing and cloud services, where tasks are generated dynamically and different tasks may become known to different processes. We measure performance in terms of the number of pending tasks, and as such it can be directly compared with the optimum number obtained under the same crash-restart-injection pattern by the best off-line algorithm. Hence, we view the problem as an online problem and we pursue competitive analysis. We propose several deterministic algorithmic solutions to the considered problem under different information models and correctness criteria, and we argue that their performance is close to the best possible offline solutions. We also prove negative results that open interesting research directions.

AB - To identify the tradeoffs between efficiency and fault-tolerance in dynamic cooperative computing, we initiate the study of a task performing problem under dynamic processes' crashes/restarts and task injections. The system consists of n message-passing processes which, subject to dynamic crashes and restarts, cooperate in performing tasks that are continuously and dynamically injected to the system. Tasks are not known a priori to the processes. This problem abstracts todays Internet-based computations, such as Grid computing and cloud services, where tasks are generated dynamically and different tasks may become known to different processes. We measure performance in terms of the number of pending tasks, and as such it can be directly compared with the optimum number obtained under the same crash-restart-injection pattern by the best off-line algorithm. Hence, we view the problem as an online problem and we pursue competitive analysis. We propose several deterministic algorithmic solutions to the considered problem under different information models and correctness criteria, and we argue that their performance is close to the best possible offline solutions. We also prove negative results that open interesting research directions.

KW - Competitive analysis

KW - Distributed algorithms

KW - Dynamic task injection

KW - Processor crashes and restarts

KW - Task execution

UR - http://www.scopus.com/inward/record.url?scp=84938914980&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84938914980&partnerID=8YFLogxK

U2 - 10.1016/j.jpdc.2015.07.007

DO - 10.1016/j.jpdc.2015.07.007

M3 - Article

AN - SCOPUS:84938914980

VL - 84

SP - 94

EP - 107

JO - Journal of Parallel and Distributed Computing

JF - Journal of Parallel and Distributed Computing

SN - 0743-7315

ER -