UsingMaintainability BasedRiskAssessment andSeverity Analysis inPrioritizing Corrective Maintenance Tasks1

A software product spends morethan65%ofits lifecycle inmaintenance. Software systems withgood maintainability canbeeasily modified tofixfaults. Wedefine maintainability-based risk asaproduct of two factors: the probability of performing maintenance tasksandtheimpactofperforming these tasks. Inthis paper, wepresent amethodology forassessing maintainability-based riskin the context ofcorrective maintenance. Theproposed methodology depends onthearchitectural artifacts andtheir evolution through thelife cycle ofthe system. Inorder toprioritize corrective maintenance tasks, we combinecomponents' maintainabilitybasedriskwiththeseverity ofafailure thatmay happen asaresult ofunfixed fault. Weillustrate the methodology onacasestudy using UML models.

[1]  John D. Musa,et al.  Software reliability engineering : more reliable software, faster development and testing , 1999 .

[2]  Hany H. Ammar,et al.  Model-based performance risk analysis , 2005, IEEE Transactions on Software Engineering.

[3]  Thomas M. Pigoski.  Practical Software Maintenance: Best Practices for Managing Your Software Investment , 1996 .

[4]  Susan A. Sherer.  Using risk analysis to manage software maintenance , 1997 .

[5]  Carolyn Seaman,et al.  Practical Software Maintenance , 2000, J. Softw. Maintenance Res. Pract..

[6]  Swapna S. Gokhale,et al.  Software defect rediscoveries: a discrete lognormal model , 2005, 16th IEEE International Symposium on Software Reliability Engineering (ISSRE'05).

[7]  Bin Li,et al.  Integrating software into PRA , 2003, 14th International Symposium on Software Reliability Engineering, 2003. ISSRE 2003..