Design and Implementation of a Contention-Aware Coscheduling Strategy on Multi-Programmed Heterogeneous Clusters

Jung-Lok YU; Hee-Jung BYUN

doi:10.1587/transinf.E94.D.2309

Design and Implementation of a Contention-Aware Coscheduling Strategy on Multi-Programmed Heterogeneous Clusters

Jung-Lok YU, Hee-Jung BYUN

Full Text Views

0

Share
Cite this

Summary :

Coscheduling has been gained a resurgence of interest as an effective technique to enhance the performance of parallel applications in multi-programmed clusters. However, existing coscheduling schemes do not adequately handle priority boost conflicts, leading to significantly degraded performance. To address this problem, in our previous study, we devised a novel algorithm that reorders the scheduling sequence of conflicting processes based on the rescheduling latency of their correspondents in remote nodes. In this paper, we exhaustively explore the design issues and implementation details of our contention-aware coscheduling scheme over Myrinet-based cluster system. We also practically analyze the impact of various system parameters and job characteristics on the performance of all considered schemes on a heterogeneous Linux cluster using a generic coscheduling framework. The results show that our approach outperforms existing schemes (by up to 36.6% in avg. job response time), reducing both boost conflict ratio and overall message delay.

Publication: IEICE TRANSACTIONS on Information Vol.E94-D No.12 pp.2309-2318

Publication Date: 2011/12/01

Publicized

Online ISSN: 1745-1361

DOI: 10.1587/transinf.E94.D.2309

Type of Manuscript: Special Section PAPER (Special Section on Parallel and Distributed Computing and Networking)

Category

Cite this

Copy

Jung-Lok YU, Hee-Jung BYUN, "Design and Implementation of a Contention-Aware Coscheduling Strategy on Multi-Programmed Heterogeneous Clusters" in IEICE TRANSACTIONS on Information, vol. E94-D, no. 12, pp. 2309-2318, December 2011, doi: 10.1587/transinf.E94.D.2309.
Abstract: Coscheduling has been gained a resurgence of interest as an effective technique to enhance the performance of parallel applications in multi-programmed clusters. However, existing coscheduling schemes do not adequately handle priority boost conflicts, leading to significantly degraded performance. To address this problem, in our previous study, we devised a novel algorithm that reorders the scheduling sequence of conflicting processes based on the rescheduling latency of their correspondents in remote nodes. In this paper, we exhaustively explore the design issues and implementation details of our contention-aware coscheduling scheme over Myrinet-based cluster system. We also practically analyze the impact of various system parameters and job characteristics on the performance of all considered schemes on a heterogeneous Linux cluster using a generic coscheduling framework. The results show that our approach outperforms existing schemes (by up to 36.6% in avg. job response time), reducing both boost conflict ratio and overall message delay.
URL: https://globals.ieice.org/en_transactions/information/10.1587/transinf.E94.D.2309/_p

Copy

@ARTICLE{e94-d_12_2309,
author={Jung-Lok YU, Hee-Jung BYUN, },
journal={IEICE TRANSACTIONS on Information},
title={Design and Implementation of a Contention-Aware Coscheduling Strategy on Multi-Programmed Heterogeneous Clusters},
year={2011},
volume={E94-D},
number={12},
pages={2309-2318},
abstract={Coscheduling has been gained a resurgence of interest as an effective technique to enhance the performance of parallel applications in multi-programmed clusters. However, existing coscheduling schemes do not adequately handle priority boost conflicts, leading to significantly degraded performance. To address this problem, in our previous study, we devised a novel algorithm that reorders the scheduling sequence of conflicting processes based on the rescheduling latency of their correspondents in remote nodes. In this paper, we exhaustively explore the design issues and implementation details of our contention-aware coscheduling scheme over Myrinet-based cluster system. We also practically analyze the impact of various system parameters and job characteristics on the performance of all considered schemes on a heterogeneous Linux cluster using a generic coscheduling framework. The results show that our approach outperforms existing schemes (by up to 36.6% in avg. job response time), reducing both boost conflict ratio and overall message delay.},
keywords={},
doi={10.1587/transinf.E94.D.2309},
ISSN={1745-1361},
month={December},}

Copy

TY - JOUR
TI - Design and Implementation of a Contention-Aware Coscheduling Strategy on Multi-Programmed Heterogeneous Clusters
T2 - IEICE TRANSACTIONS on Information
SP - 2309
EP - 2318
AU - Jung-Lok YU
AU - Hee-Jung BYUN
PY - 2011
DO - 10.1587/transinf.E94.D.2309
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E94-D
IS - 12
JA - IEICE TRANSACTIONS on Information
Y1 - December 2011
AB - Coscheduling has been gained a resurgence of interest as an effective technique to enhance the performance of parallel applications in multi-programmed clusters. However, existing coscheduling schemes do not adequately handle priority boost conflicts, leading to significantly degraded performance. To address this problem, in our previous study, we devised a novel algorithm that reorders the scheduling sequence of conflicting processes based on the rescheduling latency of their correspondents in remote nodes. In this paper, we exhaustively explore the design issues and implementation details of our contention-aware coscheduling scheme over Myrinet-based cluster system. We also practically analyze the impact of various system parameters and job characteristics on the performance of all considered schemes on a heterogeneous Linux cluster using a generic coscheduling framework. The results show that our approach outperforms existing schemes (by up to 36.6% in avg. job response time), reducing both boost conflict ratio and overall message delay.
ER -