Case study: a maintenance practice used with real-time telecommunications software

In this paper we present a case study of the software maintenance practice that has been successfully applied to real-time distributed systems, which are installed and fully operational in Moscow, St. Petersburg, and other cities across Russia. In this paper we concentrate on the software maintenance process, including customer request servicing, in-field error logging, role of information system, software deployment, and software quality policy, and especially the software quality prediction process. In this case study, the prediction process Is shown to be integral and one of the most important parts of the software maintenance process. We include a software quality prediction procedure overview and an example of the actual practice. The quality of the new software update is predicted on the basis of the current update's quantity metrics data and quality data, and new update's quantity metrics data. For management, this forecast aids software maintenance efficiency, and cost reduction. For practitioners, the most useful result presented is the process for determining the value for the break point. We end this case study with five lessons learned. Copyright (C) 2001 John Wiley & Sons, Ltd.