The Time/State-based Software-Intensive Systems Failure Mode Researches

Abstract

Nowadays the application status of Software-Intensive Systems(SISs) introduces a category of system failure caused by unforeseen operation or environment change. Generally speaking this kind of failure can be observed as system emergent behavior or degraded running. Because it relates to both the running time and state, it is called Time/State(TS)-based SISs failure. Moreover it is one of the significant sources of SISs failure. However the related researches are few. This paper presents the life cycle of software-related failure of SISs firstly. Secondly it analyzes the TS-based SISs failure mechanism and establishes the corresponding model. Moreover it introduces the traditional verification methods of SISs. Furthermore it presents the definition, classification and ontology representation of TS-based SISs failure mode. The instance validation shows the existence of TS-based SISs failure and feasibility of detecting the failure by using combined test method primarily. Finally this paper analyzes the problems and prospects the future researches.