We consider a fork-join system in which a fixed amount of computational resources has to be distributed among the K tasks forming the jobs. The queueing disciplines of the fork- and join- queues are First Come First Served. At each epoch, at most K tasks are in service while the others wait in the fork-queues. We propose an algorithm with a very simple implementation that allocates the computational resources in a way that aims at minimizing the join-queue lengths, and hence at reducing the expected job service time. We study its performance in saturation and under exponential service time and provide a methodology to derive the relevant performance indices. Explicit closed-form expressions for the expected response time and join-queue length are given for the cases of jobs consisting of two, three and four tasks.
Autori: | Sottana, Matteo (Corresponding) | |
Data di pubblicazione: | 2018 | |
Titolo: | Biased processor sharing in fork-join queues | |
Rivista: | LECTURE NOTES IN ARTIFICIAL INTELLIGENCE | |
Titolo del libro: | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | |
Digital Object Identifier (DOI): | http://dx.doi.org/10.1007/978-3-319-99154-2_17 | |
Appare nelle tipologie: | 4.1 Articolo in Atti di convegno |
File in questo prodotto:
File | Descrizione | Tipologia | Licenza | |
---|---|---|---|---|
main.pdf | Articolo principale | Documento in Post-print | Accesso chiuso-personale | Riservato |