Processor-embedded distributed smart disks for I/O-intensive workloads: Architectures, performance models and evaluation

Steve C. Chiu, Wei keng Liao, Alok N. Choudhary, Mahmut T. Kandemir

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

Processor-embedded disks, or smart disks, with their network interface controller, can in effect be viewed as processing elements with on-disk memory and secondary storage. The data sizes and access patterns of today's large I/O-intensive workloads require architectures whose processing power scales with increased storage capacity. To address this concern, we propose and evaluate disk-based distributed smart storage architectures. Based on analytically derived performance models, our evaluation with representative workloads show that offloading processing and performing point-to-point data communication improve performance over centralized architectures. Our results also demonstrate that distributed smart disk systems exhibit desirable scalability and can efficiently handle I/O-intensive workloads, such as commercial decision support database (TPC-H) queries, association rules mining, data clustering, and two-dimensional fast Fourier transform, among others.

Original languageEnglish (US)
Pages (from-to)427-446
Number of pages20
JournalJournal of Parallel and Distributed Computing
Volume64
Issue number3
DOIs
StatePublished - Mar 2004

Fingerprint

Embedded Processor
Performance Model
Workload
Performance Evaluation
Processing
Association rules
Fast Fourier transforms
Interfaces (computer)
Data mining
Scalability
Model Evaluation
Data Clustering
Association Rule Mining
Data Communication
Storage Capacity
Data storage equipment
Fast Fourier transform
Controllers
Decision Support
Communication

All Science Journal Classification (ASJC) codes

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

Cite this

@article{c9ca51d9b1564741a87b49ea158740cb,
title = "Processor-embedded distributed smart disks for I/O-intensive workloads: Architectures, performance models and evaluation",
abstract = "Processor-embedded disks, or smart disks, with their network interface controller, can in effect be viewed as processing elements with on-disk memory and secondary storage. The data sizes and access patterns of today's large I/O-intensive workloads require architectures whose processing power scales with increased storage capacity. To address this concern, we propose and evaluate disk-based distributed smart storage architectures. Based on analytically derived performance models, our evaluation with representative workloads show that offloading processing and performing point-to-point data communication improve performance over centralized architectures. Our results also demonstrate that distributed smart disk systems exhibit desirable scalability and can efficiently handle I/O-intensive workloads, such as commercial decision support database (TPC-H) queries, association rules mining, data clustering, and two-dimensional fast Fourier transform, among others.",
author = "Chiu, {Steve C.} and Liao, {Wei keng} and Choudhary, {Alok N.} and Kandemir, {Mahmut T.}",
year = "2004",
month = "3",
doi = "10.1016/j.jpdc.2004.01.004",
language = "English (US)",
volume = "64",
pages = "427--446",
journal = "Journal of Parallel and Distributed Computing",
issn = "0743-7315",
publisher = "Academic Press Inc.",
number = "3",

}

Processor-embedded distributed smart disks for I/O-intensive workloads : Architectures, performance models and evaluation. / Chiu, Steve C.; Liao, Wei keng; Choudhary, Alok N.; Kandemir, Mahmut T.

In: Journal of Parallel and Distributed Computing, Vol. 64, No. 3, 03.2004, p. 427-446.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Processor-embedded distributed smart disks for I/O-intensive workloads

T2 - Architectures, performance models and evaluation

AU - Chiu, Steve C.

AU - Liao, Wei keng

AU - Choudhary, Alok N.

AU - Kandemir, Mahmut T.

PY - 2004/3

Y1 - 2004/3

N2 - Processor-embedded disks, or smart disks, with their network interface controller, can in effect be viewed as processing elements with on-disk memory and secondary storage. The data sizes and access patterns of today's large I/O-intensive workloads require architectures whose processing power scales with increased storage capacity. To address this concern, we propose and evaluate disk-based distributed smart storage architectures. Based on analytically derived performance models, our evaluation with representative workloads show that offloading processing and performing point-to-point data communication improve performance over centralized architectures. Our results also demonstrate that distributed smart disk systems exhibit desirable scalability and can efficiently handle I/O-intensive workloads, such as commercial decision support database (TPC-H) queries, association rules mining, data clustering, and two-dimensional fast Fourier transform, among others.

AB - Processor-embedded disks, or smart disks, with their network interface controller, can in effect be viewed as processing elements with on-disk memory and secondary storage. The data sizes and access patterns of today's large I/O-intensive workloads require architectures whose processing power scales with increased storage capacity. To address this concern, we propose and evaluate disk-based distributed smart storage architectures. Based on analytically derived performance models, our evaluation with representative workloads show that offloading processing and performing point-to-point data communication improve performance over centralized architectures. Our results also demonstrate that distributed smart disk systems exhibit desirable scalability and can efficiently handle I/O-intensive workloads, such as commercial decision support database (TPC-H) queries, association rules mining, data clustering, and two-dimensional fast Fourier transform, among others.

UR - http://www.scopus.com/inward/record.url?scp=2942548209&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=2942548209&partnerID=8YFLogxK

U2 - 10.1016/j.jpdc.2004.01.004

DO - 10.1016/j.jpdc.2004.01.004

M3 - Article

AN - SCOPUS:2942548209

VL - 64

SP - 427

EP - 446

JO - Journal of Parallel and Distributed Computing

JF - Journal of Parallel and Distributed Computing

SN - 0743-7315

IS - 3

ER -