Abstract

Epidemic-style (gossip-based) techniques have recently emerged as a scalable class of protocols for peer-to-peer reliable multicast dissemination in large process groups. These protocols provide probabilistic guarantees on reliability and scalability. However, existing implementations of epidemic-style dissemination are reputed to suffer from two major drawbacks: a) (Network Overhead) when deployed on a WAN-wide or VPN-wide scale they cause a large number of packets to transit across the boundaries of network domains (eg., LANs, subnets, ASs), causing an overload on core network elements such as bridges, routers, and links, and b) (Lack of Adaptivity) they impose the same load on process group members and the network regardless of the failure characteristics of the underlying network. In this paper, we report on the first comprehensive set of solutions to these problems. The solution includes two protocols: a) a Hierarchical Gossiping protocol, and b) an Adaptive multicast Dissemination Framework that can be used with any gossiping primitive. Our solution organizes nodes in a hierarchy which re ects the network topology and guarantees low trac across domain boundaries in the network. In the interests of space, this paper focuses on the Hierarchical Gossiping protocol in detail through mathematical and simulation analysis that evaluate the improvement over a traditional gossiping protocol. We present an overview of the working and properties of the Adaptive Dissemination protocol.

Citations