Software failure rate and reliability incorporating repair policies

Abstract

Reliability of a software application, its failure rate and the residual number of faults in an application are the three most important metrics that provide a quantitative assessment of the failure characteristics of an application. Typically, one of many stochastic models known as software reliability growth models (SRGMs) is used to describe the failure behavior of an application in its testing phase, and obtain an estimate of the above metrics. In order to ensure analytical tractability, SRGMs are based on an assumption of instantaneous repair and thus the estimates of the metrics obtained using SRGMs tend to be optimistic. In practice, fault repair activity consumes a nonnegligible amount of time and resources. Also, repair may be conducted according to many policies which are reflective of the schedule and budget constraints of a project. A few research efforts that have sought to incorporate repair into SRGMs are restrictive, since they consider only one of the several SRGMs, model the repair process using a constant rate, and provide an estimate of only the residual number of faults. These techniques do not address the issue of estimating application failure rate and reliability in the presence of repair. In this paper we present a generic framework which relies on the rate-based simulation technique in order to provide the capability to incorporate various repair policies into the finite failure nonhomogeneous Poisson process (NHPP) class of software reliability growth models. We also present a technique to compute the failure rate and the reliability of an application in the presence of repair. The potential of the framework to obtain quantitative estimates of the above three metrics taking into consideration different repair policies is illustrated using several scenarios.