A feedback control mechanism for balancing I/O- and memory-intensive applications on clusters

Xiao Qin, Hong Jiang, Yifeng Zhi, David R. Swanson

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

One common assumption of existing models of load balancing is that the weights of resources and I/O buffer size are statically configured and cannot be adjusted based on a dynamic workload. Though the static configuration of these parameters performs well in a cluster where the workload can be modeled and predicted, its performance is poor in dynamic systems in which the workload is unknown. In this paper, a new feedback control mechanism is proposed to improve overall performance of a cluster with a general and practical workload including I/O-intensive and memory-intensive load. This mechanism is also shown to be effective in complementing and enhancing the performance of a number of existing dynamic load-balancing schemes. To capture the current and past workload characteristics, the primary objectives of the feedback mechanism are: (1) dynamically adjusting the resource weights, which indicate the significance of the resources, and (2) minimizing the number of page faults for memory-intensive jobs while increasing the utilization of the I/O buffers for I/O-intensive jobs by manipulating the I/O buffer size. Results from extensive trace-driven simulation experiments show that compared with a number of schemes with fixed resource weights and buffer sizes, the feedback control mechanism delivers a performance improvement in terms of the mean slowdown by up to 282% (with an average of 125%).

Original languageEnglish (US)
Pages (from-to)95-107
Number of pages13
JournalScalable Computing
Volume6
Issue number4
StatePublished - Jan 1 2005

Fingerprint

Resource allocation
Feedback control
Data storage equipment
Dynamic loads
Dynamical systems
Feedback
Experiments

Keywords

  • Cluster
  • Feedback control
  • I/O-intensive applications
  • Load balancing

ASJC Scopus subject areas

  • Computer Science(all)

Cite this

A feedback control mechanism for balancing I/O- and memory-intensive applications on clusters. / Qin, Xiao; Jiang, Hong; Zhi, Yifeng; Swanson, David R.

In: Scalable Computing, Vol. 6, No. 4, 01.01.2005, p. 95-107.

Research output: Contribution to journalArticle

@article{50ea00411d7d4f0e8542f017177d34f4,
title = "A feedback control mechanism for balancing I/O- and memory-intensive applications on clusters",
abstract = "One common assumption of existing models of load balancing is that the weights of resources and I/O buffer size are statically configured and cannot be adjusted based on a dynamic workload. Though the static configuration of these parameters performs well in a cluster where the workload can be modeled and predicted, its performance is poor in dynamic systems in which the workload is unknown. In this paper, a new feedback control mechanism is proposed to improve overall performance of a cluster with a general and practical workload including I/O-intensive and memory-intensive load. This mechanism is also shown to be effective in complementing and enhancing the performance of a number of existing dynamic load-balancing schemes. To capture the current and past workload characteristics, the primary objectives of the feedback mechanism are: (1) dynamically adjusting the resource weights, which indicate the significance of the resources, and (2) minimizing the number of page faults for memory-intensive jobs while increasing the utilization of the I/O buffers for I/O-intensive jobs by manipulating the I/O buffer size. Results from extensive trace-driven simulation experiments show that compared with a number of schemes with fixed resource weights and buffer sizes, the feedback control mechanism delivers a performance improvement in terms of the mean slowdown by up to 282{\%} (with an average of 125{\%}).",
keywords = "Cluster, Feedback control, I/O-intensive applications, Load balancing",
author = "Xiao Qin and Hong Jiang and Yifeng Zhi and Swanson, {David R.}",
year = "2005",
month = "1",
day = "1",
language = "English (US)",
volume = "6",
pages = "95--107",
journal = "Scalable Computing",
issn = "1895-1767",
publisher = "universitatea de vest",
number = "4",

}

TY - JOUR

T1 - A feedback control mechanism for balancing I/O- and memory-intensive applications on clusters

AU - Qin, Xiao

AU - Jiang, Hong

AU - Zhi, Yifeng

AU - Swanson, David R.

PY - 2005/1/1

Y1 - 2005/1/1

N2 - One common assumption of existing models of load balancing is that the weights of resources and I/O buffer size are statically configured and cannot be adjusted based on a dynamic workload. Though the static configuration of these parameters performs well in a cluster where the workload can be modeled and predicted, its performance is poor in dynamic systems in which the workload is unknown. In this paper, a new feedback control mechanism is proposed to improve overall performance of a cluster with a general and practical workload including I/O-intensive and memory-intensive load. This mechanism is also shown to be effective in complementing and enhancing the performance of a number of existing dynamic load-balancing schemes. To capture the current and past workload characteristics, the primary objectives of the feedback mechanism are: (1) dynamically adjusting the resource weights, which indicate the significance of the resources, and (2) minimizing the number of page faults for memory-intensive jobs while increasing the utilization of the I/O buffers for I/O-intensive jobs by manipulating the I/O buffer size. Results from extensive trace-driven simulation experiments show that compared with a number of schemes with fixed resource weights and buffer sizes, the feedback control mechanism delivers a performance improvement in terms of the mean slowdown by up to 282% (with an average of 125%).

AB - One common assumption of existing models of load balancing is that the weights of resources and I/O buffer size are statically configured and cannot be adjusted based on a dynamic workload. Though the static configuration of these parameters performs well in a cluster where the workload can be modeled and predicted, its performance is poor in dynamic systems in which the workload is unknown. In this paper, a new feedback control mechanism is proposed to improve overall performance of a cluster with a general and practical workload including I/O-intensive and memory-intensive load. This mechanism is also shown to be effective in complementing and enhancing the performance of a number of existing dynamic load-balancing schemes. To capture the current and past workload characteristics, the primary objectives of the feedback mechanism are: (1) dynamically adjusting the resource weights, which indicate the significance of the resources, and (2) minimizing the number of page faults for memory-intensive jobs while increasing the utilization of the I/O buffers for I/O-intensive jobs by manipulating the I/O buffer size. Results from extensive trace-driven simulation experiments show that compared with a number of schemes with fixed resource weights and buffer sizes, the feedback control mechanism delivers a performance improvement in terms of the mean slowdown by up to 282% (with an average of 125%).

KW - Cluster

KW - Feedback control

KW - I/O-intensive applications

KW - Load balancing

UR - http://www.scopus.com/inward/record.url?scp=67649540227&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=67649540227&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:67649540227

VL - 6

SP - 95

EP - 107

JO - Scalable Computing

JF - Scalable Computing

SN - 1895-1767

IS - 4

ER -