Authors:
Gregory V. Chockler,
Nabil Huleihel and
Danny Dolev.
In Proceedings of the Seventeenth ACM Symposium on Principles of
Distributed Computing (PODC'98), June 28 - July 2 1998,
Puerto Vallarta, Mexico.
Abstract:
In this paper we present a novel algorithm that implements a totally
ordered multicast primitive for a Totally Ordered Group
Communication Service (TO-GCS). TO-GCS is a powerful infrastructure
for building distributed fault-tolerant applications, such as
totally ordered broadcast, consistent object replication, distributed
shared memory, Computer Supported Cooperative Work (CSCW) applications
and distributed monitoring and display applications.
Our algorithm is adaptive, i.e., it is able to dynamically alter
the message delivery order in response to changes in the transmission
rates of the participating processes. This compensates for differences
among participant transmission rates and therefore minimizes
fluctuations in message delivery latency. Our algorithm is thus
useful for soft real-time environments where sharp fluctuations in
message delivery latency are not acceptable.
Our solution provides well-defined message ordering semantics. These
semantics are preserved even in the face of site and communication
link failures.
Postscript Version:
ps,
ps.gz.
Full version available as Technical Report CS98-4, Institute of
Computer Science, The Hebrew University of Jerusalem, Jerusalem,
Israel, May 1998:
ps,
ps.gz,
abstract.
grishac@cs.huji.ac.il
Last modified: Thu Apr 9 13:09:48 IDT 1998