On optimal hierarchical configuration of distributed systems on mesh and hypercube

Dajin Wang, Jiannong Cao

Research output: Chapter in book / Conference proceedingConference article published in proceeding or bookAcademic researchpeer-review

3 Citations (Scopus)

Abstract

This paper studies hierarchical configuration of distributed systems for achieving optimized system performance. A distributed system consists of a collection of local processes which are distributed over the network of processors and cooperate to perform some functions. An hierarchical approach is to group and organize the distributed processes into a logical hierarchy of multiple levels, so as to coordinate the local computation/control activities to improve the overall system performance. It has been proposed as an effective way to solve various problems in distributed computing, such as distributed monitoring, resource scheduling, and network routing. The optimization problem considered in this paper is concerned with finding an optimal hierarchical partition of the processors so that the total cost is minimal. The problem in its general form has been known to be NP-hard. Therefore, we just focus on distributed computing jobs which require collecting information from all processors. One example of such job is distributed monitoring. By limiting the levels of the hierarchy to 2, we study optimal hierarchical configurations for two popular interconnection networks: mesh and hypercube. Based on analytical results, partitioning algorithms are proposed which achieve optimal communication cost attributed to information collection. We also discuss heuristic schemes for multiple-level hierarchical partitions. Although this paper is presented in terms of distributed monitoring, the approaches can also be applied to other hierarchical control problems in distributed computing.
Original languageEnglish
Title of host publicationProceedings - International Parallel and Distributed Processing Symposium, IPDPS 2003
PublisherIEEE
ISBN (Electronic)0769519261, 9780769519265
DOIs
Publication statusPublished - 1 Jan 2003
EventInternational Parallel and Distributed Processing Symposium, IPDPS 2003 - Nice, France
Duration: 22 Apr 200326 Apr 2003

Conference

ConferenceInternational Parallel and Distributed Processing Symposium, IPDPS 2003
CountryFrance
CityNice
Period22/04/0326/04/03

ASJC Scopus subject areas

  • Computational Theory and Mathematics
  • Theoretical Computer Science
  • Software

Cite this