Error confinement/data recovery in distributed systems
Manel Ben-Romdhane, Bernard Courtois, Centre National de la Recherche Scientifique, LA 7, Laboratoire d'Informatique et de Mathematiques Appliquees de Grenoble Grenoble (FRANCE), Universite de Grenoble 1, Institut National Polytechnique de Grenoble Grenoble (FRANCE) · OpenGrey (Institut de l'Information Scientifique et Technique) · 1982
The aim of this paper is to take advantage of distributed systems for fault-tolerance, but keeping in mind that some advantages may be greatly diminished by delays in detection of failures and/or to the transmission of information. The failures which can be considered may be due to hardware or software. The method used is to record some information in a buffer before using it, and the study is concerned with the method of determining when the information can be used. The necessity of using such a mechanism in order to be able to recover duplicated data is illustrated in particular for the nonstop operation of a telephone system.