Modeling and performance analysis to predict the behavior of a divisible load application in a star network cloud

Leila Ismail, Liren Zhang

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

1 Citation (Scopus)

Abstract

In this work, we introduce a model to predict the behavior of a divisible load application using a producerconsumer pattern, also called master-computing worker pattern when running in a Star network Cloud. Within the Cloud, we consider 3 main components: the data transmission capacity, the data receiving capacity and the computing capacity of the Cloud. The objective of this work is to help software engineers to analyze and predict the behavior of their divisible load applications before the implementation phase considering Cloud capacities. In particular, taking into account the effects of networking and processing on the performance of the Cloud, we answer the following questions: 1) How can the efficiency of a computing worker be measured?, 2) What is the average processing time for a task in a computing worker?. Finally, we evaluate the proposed model by conducting a numerical analysis.

Original languageEnglish
Title of host publicationProceedings - UKSim 4th European Modelling Symposium on Computer Modelling and Simulation, EMS2010
Pages369-374
Number of pages6
DOIs
Publication statusPublished - Dec 1 2010
EventUKSim 4th European Modelling Symposium on Computer Modelling and Simulation, EMS2010 - Pisa, Italy
Duration: Nov 17 2010Nov 19 2010

Publication series

NameProceedings - UKSim 4th European Modelling Symposium on Computer Modelling and Simulation, EMS2010

Other

OtherUKSim 4th European Modelling Symposium on Computer Modelling and Simulation, EMS2010
Country/TerritoryItaly
CityPisa
Period11/17/1011/19/10

Keywords

  • Cloud computing
  • Distributed systems
  • Divisible load application
  • Performance analysis

ASJC Scopus subject areas

  • Applied Mathematics
  • Modelling and Simulation

Fingerprint

Dive into the research topics of 'Modeling and performance analysis to predict the behavior of a divisible load application in a star network cloud'. Together they form a unique fingerprint.

Cite this