Conceptual Models for the Reliability of Diverse Systems - New Results

by Peter Popov and Lorenzo Strigini presented at the 28-th annual Symposium on Fault-Tolerant Computing (FTCS'28), Munich, Germany, June 1998. The full text of the paper is available for download in .pdf format. The abstract is given below. 

Abstract

We address problems in modelling the reliability of multiple-version software, and present models intended to improve the understanding of the various ways failure dependence between versions can arise. The previous models, by Eckhardt and Lee and by Littlewood and Miller, described what behaviour could be expected "on average" from a randomly chosen pair of "independently generated" versions. Instead, we address the problem of predicting the reliability of a specific pair of versions. The concept of "variation of difficulty" between situations to which software may be subject is central to the previous models cited. We show that it has even more far-reaching implications than previously found. In particular, we consider the practical implications of two phenomena: varying probabilities of failure over input sub-domains or operating regimes; and positive correlation between successive executions of control software. Our analysis provides some practical advice for regulators, and useful insight into non-intuitive aspects of the failure process of diverse software.


Estimating Bounds on the Reliability of Diverse Systems

by Peter Popov, Lorenzo Strigini, John May and Silke Kuball is a substantially enhanced version of the FTCS-98 paper, "Estimating Bounds on the Reliability of Diverse Systems". It has been accepted for publication in IEEE Transactions on Software Engineering. The full text in .pdf is accessible provided the IEEE copyright retrictions are respected. The abstract is given below.

Abstract

We address the problems of estimating the reliability of multiple-version software, and improve the understanding of the various ways failure dependence between versions can arise. Specifically, we step from the previous conceptual models, which described what behaviour could be expected "on average" from a randomly chosen pair of "independently generated" versions to predictions using specific information about a given pair of versions. The concept of "variation of difficulty" between situations to which software may be subject is central to the previous models cited. We show that it has more far-reaching implications than previously found.
We show the practical implications of varying probabilities of failure over input subdomains or operating regimes. A direct practical gain for designers, users and regulators is the possibility of estimating upper and lower bounds on the reliability of a two-versions system. In certain circumstances these will be narrower than could be estimated with other methods.

Page maintained by: Peter Popov
Last modified 3 July 2002.