[comp.sys.isis] Reminder: Tutorial in Houston

ken@gvax.cs.cornell.edu (Ken Birman) (10/19/90)

________________________________________________________________________

         TUTORIAL ON REPLICATION IN DISTRIBUTED SYSTEMS

                The Houstonian Hotel, Houston, TX
                        November 7, 1990

In conjunction with the first IEEE Workshop on the Management of
Replicated Data, the IEEE Computer Society technical committee on
operating systems and application environments (TCOS) organizes
on November 7 a tutorial on Replication in Distributed Systems by
Kenneth Birman from Cornell University.

OVERVIEW

Replication is sometimes viewed as primarily useful in building
fault-tolerant distributed systems.  Yet, replication underlies a
wide range of distributed functions, and may be as fundamental a
distributed computing technology as is communication.  This six-
hour tutorial will focus on uses of replication in distributed
systems, through a series of case-studies.  We will review the
best known replication techniques and discuss replication in sys-
tems such as the ISIS Toolkit, the Linda tuple space, and IBM's
HAS.  Other topics will include distributed consistency, transac-
tional replication, virtual synchrony, realtime issues, and
replication in NUMA multiprocessor settings.

The tutorial will focus on practical issues rather than theory,
and assumes a solid background in networking (telecommunications)
but not necessarily distributed computing.  Lecture notes, refer-
ences to supplemental readings, and copies of slides will be pro-
vided to attendees free of charge.

THE LECTURER

Kenneth P. Birman is Associate Professor in the Department of
Computer Science at Cornell University, and President of ISIS
Distributed Systems, a consulting firm specializing in distri-
buted computing technologies.  His ISIS Project has produced a
well known and widely used toolkit for constructing distributed
and fault-tolerant software, and he has published extensively on
replication and fault-tolerance.  His current research is con-
cerned with large-scale distributed applications and distributed
control problems.

_________________________________________________________________

Name:             ________________________________________________
Organization:     ________________________________________________
Address:          ________________________________________________
City and State:   ________________________________________________
Telephone:        ________________________________________________

Fees                   Before October 1            On-site
Tutorial and lunch           $150                    $200
Lodging                $78/night single or double occupancy

Payment must be made in U.S. funds by a check made payable to
``Workshop on the Management of Replicated Data'' drawn on a U.S.
bank.  All hotel arrangements should be made directly with the
Houstonian Hotel: 1-800-231-2759 or 1-800-392-0784 (in Texas)

Return this form to:   Luis Felipe Cabrera,
                       IBM Almaden Research Center
                       Mail-code K55/803
                       650 Harry Road,
                       San Jose, CA 95120-6099.