To migrate or to wait: Bandwidth-latency tradeoff in opportunistic scheduling of parallel tasks

Ting He, Shiyao Chen, Hyoil Kim, Lang Tong, Kang Won Lee

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

5 Scopus citations

Abstract

We consider the problem of scheduling low-priority tasks onto resources already assigned to high-priority tasks. Due to burstiness of the high-priority workloads, the resources can be temporarily underutilized and made available to the low-priority tasks. The increased level of utilization comes at a cost to the low-priority tasks due to intermittent resource availability. Focusing on two major costs, bandwidth cost associated with migrating tasks and latency cost associated with suspending tasks, we aim at developing online scheduling policies achieving the optimal bandwidth-latency tradeoff for parallel low-priority tasks with synchronization requirements. Under Markovian resource availability models, we formulate the problem as a Markov Decision Process (MDP) whose solution gives the optimal scheduling policy. Furthermore, we discover structures of the problem in the special case of homogeneous availability patterns that enable a simple threshold-based policy that is provably optimal. We validate the efficacy of the proposed policies by trace-driven simulations.

Original languageEnglish (US)
Title of host publication2012 Proceedings IEEE INFOCOM, INFOCOM 2012
Pages2871-2875
Number of pages5
DOIs
StatePublished - 2012
EventIEEE Conference on Computer Communications, INFOCOM 2012 - Orlando, FL, United States
Duration: Mar 25 2012Mar 30 2012

Publication series

NameProceedings - IEEE INFOCOM
ISSN (Print)0743-166X

Other

OtherIEEE Conference on Computer Communications, INFOCOM 2012
CountryUnited States
CityOrlando, FL
Period3/25/123/30/12

All Science Journal Classification (ASJC) codes

  • Computer Science(all)
  • Electrical and Electronic Engineering

Fingerprint Dive into the research topics of 'To migrate or to wait: Bandwidth-latency tradeoff in opportunistic scheduling of parallel tasks'. Together they form a unique fingerprint.

Cite this