ORMC'96 Invited Talk:
Bayou: A Replicated Storage System for Mobile Environments
Abstract
Bayou is a replicated, weakly consistent storage system designed for a
mobile computing environment that includes portable machines with less
than ideal network connectivity. To maximize availability, users can read
from and write to any accessible replica. Bayou's design has focused on
supporting application-specific mechanisms to detect and resolve the update
conflicts that naturally arise in such a system, ensuring that replicas
move towards eventual consistency, and defining a protocol by which the
resolution of update conflicts stabilizes. It includes novel methods for
conflict detection, called dependency checks, and per-write conflict
resolution based on client-provided merge procedures.
Bayou's design also exhibits good scalability and adaptability since no
operation involves more than two machines:
- Clients can read from and write to one server at the time.
- Servers propagate writes among themselves via a pairwise anti-entropy
protocol that permits incremental exchanges of updates.
- A new database replica, i.e. server, can be created from any existing
replica.
In the talk I will describe Bayou's architecture, discuss some of
the implementation and security issues, as well our plans to explore
Bayou as an infrastructure for asynchronous collaborative applications.
Brief Bio
Karin Petersen has been a Member of the Research Staff at Xerox PARC's
Computer Science Laboratory since September of 1993. Prior to joining
PARC, Karin received an Engineering degree in Computer Science from the
Simon Bolivar University (Venezuela) in 1988, and a Ph.D. in Computer
Science from Princeton University in 1993. Her research focuses on
distributed and parallel systems, performance evaluation, and user
interfaces for mobile and distributed applications.
Last modified: 18.04.2002 (Original version last modified prior to Sun Oct 6, 1996)