Software Failure Investigation : A Near-Miss Analysis Approach / by Jan Eloff, Madeleine Bihina Bella.

Chap1: Introduction -- Chap2: Software Failures: An Overview -- Chap3: Near-Miss Analysis: An Overview -- Chap4: A Methodology for Investigating Software Failures Using Digital Forensics and Near-Miss Analysis -- Chap5: A Well-Defined Model for Near-Miss Detection And Prioritisation -- Chap6: An Architecture for a Near-Miss System (Nms) -- Chap7: Practical Demonstration of Conducting a Near-Miss Analysis Investigation for Software Failures.

This book reviews existing operational software failure analysis techniques and proposes near-miss analysis as a novel, and new technique for investigating and preventing software failures. The authors provide details on how near-miss analysis techniques focus on the time-window before the software failure actually unfolds, so as to detect the high-risk conditions that can lead to a major failure. They detail how by alerting system users of an upcoming software failure, the detection of near misses provides an opportunity to collect at runtime failure-related data that is complete and relevant. They present a near-miss management systems (NMS) for detecting upcoming software failures, which can contribute significantly to the improvement of the accuracy of the software failure analysis. A prototype of the NMS is implemented and is discussed in the book. The authors give a practical hands-on approach towards doing software failure investigations by means of near-miss analysis that is of use to industry and academia. Investigates software failures in an operational environment through near-miss analysis; Studies recent and major software failures of significant magnitude; Provides a design of an architecture for a Near-Miss Management System.