Fault/failure analysis of the Tandem NonStop-UX operating system

Abstract

This paper presents the results of an analyse's of failures in several releases of Tandem's NonStop-UX operating system. NonStop-UX is based on UNIX System V. The analysis covers software failures from the field and failures reported by Tandem's test center. Faults are classified based on the status of the reported failures, the locations of the code that detected the problems, the panic messages generated by the systems, the faulty source modules, and the types of developer's mistakes. We present distributions of the failure and repair times for unique and duplicate failures. We also discuss how the analysis results can be used for assessing the dependability of the operating system and guiding improvement efforts.