Data-Aware Multicast

  • Authors:
  • Sebastien Baehni;Patrick Th. Eugster;Rachid Guerraoui

  • Affiliations:
  • Distributed Programming Laboratory, EPFL;Distributed Programming Laboratory, EPFL;Distributed Programming Laboratory, EPFL

  • Venue:
  • DSN '04 Proceedings of the 2004 International Conference on Dependable Systems and Networks
  • Year:
  • 2004

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper presents a multicast algorithm for peer-to-peerdissemination of events in a distributed topic-basedpublish-subscribe system, where processes publish events ofcertain topics, organized in a hierarchy, and expect eventsof topics they subscribed to. Our algorithm is "data-aware"in the sense that it exploits information about process subscriptionsand topic inclusion relationships to build dynamicgroups of processes and efficiently manage the flow ofinformation within and between these process groups. This"data-awareness" helps limit the membership informationthat each process needs to maintain and preserves processesfrom receiving messages related to topics they have not subscribedto. It also provides the application with means tocontrol, for each topic in a hierarchy, the trade-off betweenthe message complexity and the reliability of event dissemination.We convey this trade-off through both analysis andsimulation.