Resource management for elastic publish subscribe systems: A performance modeling-based approach

Phuong Nguyen, Klara Nahrstedt

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

Abstract

As more and more information systems are moving to the cloud, there have been efforts to deploy publishsubscribe (or pub/sub) systems in the cloud environment to take advantage of the elasticity of resources. As a result, there is a need to perform resource management for the cloudbased pub/sub systems that support various types of jobs, each consists of a series of tasks, or a workflow. Designing an efficient and effective resource management approach for the cloudbased pub/sub system is challenging because such an approach needs to be able to support flexible provisioning strategies, model the complex interactions between heterogeneous types of jobs, and provide dynamic resource allocation capability. In this paper, we formulate the resource management problem of elastic pub/sub system as optimization problems using different objectives functions. We model the elastic pub/sub system as a multiple-class open queuing network to derive system performance measures, and propose greedy algorithms to efficiently solve the optimization problems. Our evaluation based on simulation on real system show that our proposed solution outperforms the baseline and is robust in dealing with high-volume and fast-changing workload.

Original languageEnglish (US)
Title of host publicationProceedings - 2016 IEEE 9th International Conference on Cloud Computing, CLOUD 2016
EditorsIan Foster, Nimish Radia, Ian Foster
PublisherIEEE Computer Society
Pages561-568
Number of pages8
ISBN (Electronic)9781509026197
DOIs
StatePublished - Jan 17 2017
Event9th International Conference on Cloud Computing, CLOUD 2016 - San Francisco, United States
Duration: Jun 27 2016Jul 2 2016

Publication series

NameIEEE International Conference on Cloud Computing, CLOUD
ISSN (Print)2159-6182
ISSN (Electronic)2159-6190

Other

Other9th International Conference on Cloud Computing, CLOUD 2016
CountryUnited States
CitySan Francisco
Period6/27/167/2/16

ASJC Scopus subject areas

  • Artificial Intelligence
  • Information Systems
  • Software

Fingerprint Dive into the research topics of 'Resource management for elastic publish subscribe systems: A performance modeling-based approach'. Together they form a unique fingerprint.

Cite this