Forming SPN-MapReduce Model for Estimation Job Execution Time in Cloud Computing

Ying Jun Chen, Gwo Jiun Horng, Sheng-Tzong Cheng, His Chuan Wang

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

MapReduce is a key technique among several new concepts in cloud computing. When running a MapReduce job, programmers cannot obtain information about how the performance of their application will be in their own test environment. Several parameters affect the performance of MapReduce, programmers must spend a substantial amount of time identifying the most suitable parameters or studying execution details in MapReduce. Execution details in MapReduce are examined in depth and described using Stochastic Petri Net (SPN) to then develop the SPN-MapReduce (SPN-MR) model. To analyze the performance of SPN-MR, mean delay time formulas are defined for each timed transition. SPN-MR is the first proposed which can estimate the execution time of MapReduce jobs with a known input data size in hundreds of milliseconds and reduce the time spent by programmers in tuning performance. The experimental results of SPN-MR are compared with two benchmarks of actual tests and the average error range is found to be within 5 % under the size of 10 GB input data. Therefore, SPN-MR can enable MapReduce programmers to evaluate performance effectively and help programmers understand execution details in MapReduce.

Original languageEnglish
Pages (from-to)3465-3493
Number of pages29
JournalWireless Personal Communications
Volume94
Issue number4
DOIs
Publication statusPublished - 2017 Jun 1

Fingerprint

Cloud computing
Petri nets
Time delay
Tuning

All Science Journal Classification (ASJC) codes

  • Computer Science Applications
  • Electrical and Electronic Engineering

Cite this

Chen, Ying Jun ; Horng, Gwo Jiun ; Cheng, Sheng-Tzong ; Wang, His Chuan. / Forming SPN-MapReduce Model for Estimation Job Execution Time in Cloud Computing. In: Wireless Personal Communications. 2017 ; Vol. 94, No. 4. pp. 3465-3493.
@article{9f4b453890cd4209a949ad7122ef0ffb,
title = "Forming SPN-MapReduce Model for Estimation Job Execution Time in Cloud Computing",
abstract = "MapReduce is a key technique among several new concepts in cloud computing. When running a MapReduce job, programmers cannot obtain information about how the performance of their application will be in their own test environment. Several parameters affect the performance of MapReduce, programmers must spend a substantial amount of time identifying the most suitable parameters or studying execution details in MapReduce. Execution details in MapReduce are examined in depth and described using Stochastic Petri Net (SPN) to then develop the SPN-MapReduce (SPN-MR) model. To analyze the performance of SPN-MR, mean delay time formulas are defined for each timed transition. SPN-MR is the first proposed which can estimate the execution time of MapReduce jobs with a known input data size in hundreds of milliseconds and reduce the time spent by programmers in tuning performance. The experimental results of SPN-MR are compared with two benchmarks of actual tests and the average error range is found to be within 5 {\%} under the size of 10 GB input data. Therefore, SPN-MR can enable MapReduce programmers to evaluate performance effectively and help programmers understand execution details in MapReduce.",
author = "Chen, {Ying Jun} and Horng, {Gwo Jiun} and Sheng-Tzong Cheng and Wang, {His Chuan}",
year = "2017",
month = "6",
day = "1",
doi = "10.1007/s11277-016-3786-7",
language = "English",
volume = "94",
pages = "3465--3493",
journal = "Wireless Personal Communications",
issn = "0929-6212",
publisher = "Springer Netherlands",
number = "4",

}

Forming SPN-MapReduce Model for Estimation Job Execution Time in Cloud Computing. / Chen, Ying Jun; Horng, Gwo Jiun; Cheng, Sheng-Tzong; Wang, His Chuan.

In: Wireless Personal Communications, Vol. 94, No. 4, 01.06.2017, p. 3465-3493.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Forming SPN-MapReduce Model for Estimation Job Execution Time in Cloud Computing

AU - Chen, Ying Jun

AU - Horng, Gwo Jiun

AU - Cheng, Sheng-Tzong

AU - Wang, His Chuan

PY - 2017/6/1

Y1 - 2017/6/1

N2 - MapReduce is a key technique among several new concepts in cloud computing. When running a MapReduce job, programmers cannot obtain information about how the performance of their application will be in their own test environment. Several parameters affect the performance of MapReduce, programmers must spend a substantial amount of time identifying the most suitable parameters or studying execution details in MapReduce. Execution details in MapReduce are examined in depth and described using Stochastic Petri Net (SPN) to then develop the SPN-MapReduce (SPN-MR) model. To analyze the performance of SPN-MR, mean delay time formulas are defined for each timed transition. SPN-MR is the first proposed which can estimate the execution time of MapReduce jobs with a known input data size in hundreds of milliseconds and reduce the time spent by programmers in tuning performance. The experimental results of SPN-MR are compared with two benchmarks of actual tests and the average error range is found to be within 5 % under the size of 10 GB input data. Therefore, SPN-MR can enable MapReduce programmers to evaluate performance effectively and help programmers understand execution details in MapReduce.

AB - MapReduce is a key technique among several new concepts in cloud computing. When running a MapReduce job, programmers cannot obtain information about how the performance of their application will be in their own test environment. Several parameters affect the performance of MapReduce, programmers must spend a substantial amount of time identifying the most suitable parameters or studying execution details in MapReduce. Execution details in MapReduce are examined in depth and described using Stochastic Petri Net (SPN) to then develop the SPN-MapReduce (SPN-MR) model. To analyze the performance of SPN-MR, mean delay time formulas are defined for each timed transition. SPN-MR is the first proposed which can estimate the execution time of MapReduce jobs with a known input data size in hundreds of milliseconds and reduce the time spent by programmers in tuning performance. The experimental results of SPN-MR are compared with two benchmarks of actual tests and the average error range is found to be within 5 % under the size of 10 GB input data. Therefore, SPN-MR can enable MapReduce programmers to evaluate performance effectively and help programmers understand execution details in MapReduce.

UR - http://www.scopus.com/inward/record.url?scp=84989932445&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84989932445&partnerID=8YFLogxK

U2 - 10.1007/s11277-016-3786-7

DO - 10.1007/s11277-016-3786-7

M3 - Article

AN - SCOPUS:84989932445

VL - 94

SP - 3465

EP - 3493

JO - Wireless Personal Communications

JF - Wireless Personal Communications

SN - 0929-6212

IS - 4

ER -