Relationship between software availability measurement and the number of restorations with imperfect debugging

In this paper, we propose a software availability model considering the number of restoration actions. We correlate the failure and restoration characteristics of the software system with the cumulative number of corrected faults. Furthermore, we consider an imperfect debugging environment where the detected faults are not always corrected and removed from the system. The time-dependent behavior of the system alternating between up and down states is described by a Markov process. From this model, we can derive quantitative measures for software availability assessment based on the number of restoration actions. Finally, we show numerical examples of software availability analysis.