Contents   Index

Preface

This is the Master Thesis of Michael G. Sørensen. It was written as part of the AMIGOS¹ (Advanced Mobile Integration in General Operating Systems) project at DIKU, Department of Computer Science, University of Copenhagen. The subproject was nicknamed $\mathcal{P}$$e$$\mathcal{S}$$t$$\mathcal{O}$ for PEssimistic, STrict, and Optimistic.

Prerequisites: The reader is assumed to have knowledge of distributed operating systems corresponding to having read and understood [52, Ch.1-6], e.g., by following the course "Distributed Operating Systems" at DIKU.

The basic ideas of the presented model for consistency stem from the report "Transactions in Mobile Computing" [34], which I recommend is read beforehand, as much of the material in it will be referred to rather than repeated in this report.

A Model for Multi-Level Consistency: During the writing phase of this report, I wrote an article [51] and submitted it for the OOPSLA'96 Workshop on Object Replication and Mobile Computing (ORMC'96). It was accepted and therefore I attended the workshop (in San José, California on the 7. November 1996) and did a presentation. I consider the writing of the article and the presentation as a part of my Master Thesis, and it is therefore the reader will find a copy of the article "attached".

This report (DIKU 96-3-7) is available for download in gzip'ed postscript format: $<$ftp://ftp.diku.dk/diku/distlab/amigos/diku-96-3-7.ps.gz$>$. And the attached article as: $<$ftp://ftp.diku.dk/diku/distlab/ormc96/o12.ps.gz$>$.

All the systems mentioned herein are trademarks of their respective companies and owners.

Acknowledgements: Birger Andersen for pushing me forward, but not over the edge! Jørgen Sværke Hansen & Torben Reich for letting me "steal" a number of items from their TACO Project.²

BIG THANKS GO TO
Christina
for her enormous patience

Footnotes

... AMIGOS¹

AMIGOS: $<$http://www.diku.dk/distlab/amigos/$>$

... Project.²

TACO: $<$http://www.diku.dk/distlab/amigos/taco.html$>$

  Contents   Index