Lookup NU author(s): Professor Alexander Romanovsky
Full text for this publication is not currently held within this repository. Alternative links are provided below where available.
The purpose of the paper is to present a recovery block (RB) scheme that is suitable for a real time application. For this purpose, it has to have a predictable fault tolerant behaviour. We analyse the basic problems to be tackled for introducing the RB scheme into real time systems and propose some approaches and solutions allowing to handle them. Steps of calculating the RB worst case execution time are discussed; in particular, several approaches to reducing this time and making it easier to estimate more precisely are considered. We address the problems of the RB implementation and the original language construct extensions which allow to guarantee the predictability of the RB behaviour in the run time. In particular, features which guarantee the fault tolerance of the RB scheme and, as a result, allow to predict its behaviour in cases of software and hardware faults are discussed. The differences between using RBs for hard real time tasks and for soft ones are considered. The computational and timing model of the RB execution is described. These RBs can be used as building blocks (parts of tasks) for designing real time systems with predictable behaviour.
Author(s): Romanovsky AB
Publication type: Article
Publication status: Published
Journal: Computer Systems Science and Engineering
Print publication date: 01/11/1998
ISSN (print): 0267-6192
Publisher: CRL Publishing Ltd.