A research project to investigate the design and construction of reliable computing systems was initiated by B. Randell at the University of Newcastle upon Tyne in 1972. In over ten years of research on system reliability, a substantial number of papers have been produced by the members of this project. These papers have appeared in a variety of journals and conference proceedings and it is hoped that this book will prove to be a convenient reference volume for research workers active in this important area. In selecting papers published by past and present members of this project, I have used the following criteria: a paper is selected if it is concerned with fault tolerance and is not a review paper and was published before 1983. I have used these criteria (with only one or two exceptions!) in order to present a collection of papers with a common theme and, at the same time, to limit the size of the book to a reasonable length. The papers have been grouped into seven chapters. The first chapter introduces fundamental concepts of fault tolerance and ends with the earliest Newcastle paper on reliability. The project perhaps became well known after the invention of recovery blocks - a simple yet effective means of incorporating fault tolerance in software. The second chapter contains papers on recovery blocks, starting with the paper which first introduced the concept.

Inhalt

1. System Reliability.- Software Fault Tolerance Terminology Proposals.- System Structure for Software Fault Tolerance.- Operating Systems: The Problems of Performance and Reliability.- 2. Recovery Blocks.- A Program Structure for Error Detection and Recovery.- A Reconsideration of the Recovery Block Scheme.- Recovery Blocks in Action: A System Supporting High Reliability.- Sequential Pascal with Recovery Blocks.- Fault Tolerant Sequential Programming Using Recovery Blocks.- A Recovery Cache for the PDP-11.- Recovery and Crash Resistance in a Filing System.- 3. Exception Handling.- Software Reliability: The Role of Programmed Exception Handling.- Exception Handling and Software Fault Tolerance.- Robust Data Types.- Systematic Detection of Exception Occurrences.- Safe Programming.- 4. Concurrent Systems.- Process Structuring, Synchronisation and Recovery Using Atomic Actions.- A Formal Model of Atomicity in Asynchronous Systems.- Reliable Resource Allocation Between Unreliable Processes.- Concurrent Pascal with Backward Error Recovery: Language Features and Examples.- Concurrent Pascal with Backward Error Recovery: Implementation.- A Framework for Software Fault Tolerance in Real-Time Systems.- 5. Multilevel Systems.- A Model of Recoverability in Multilevel Systems.- The Provision of Recoverable Interfaces.- Structuring Distributed Systems for Recoverability and Crash Resistance.- 6. Distributed Systems.- Consistent State Restoration in Distributed Systems.- Recovery Control of Communicating Processes in a Distributed System.- A Dependency Commitment and Recovery Model for Atomic Actions.- Fail-Safe Extrema-Finding in a Circular Distributed System.- The Design of a Reliable Remote Procedure Call Mechanism.- Reliable Remote Calls for Distributed Unix: An Implementation Study.- The Newcastle Connection or UNIXes of the World Unite.- Recoverability Aspects of a Distributed File System.- 7. Fault Tolerance and System Structuring.- Fault Tolerance and System Structuring.