Performance analysis of software maintenance process using stochastic Petri nets

Software Maintenance is a time taking activity in the real world. Execution time of software maintenance process may get increased due to interdepartmental communication, thus increasing the cost and decreasing the performance of the process. We suggested performance evaluation of software maintenance process through the transformation of activity diagram of software maintenance process into GSPN. It evaluates the performance of the maintenance process before actual process is being executed, unlike other, so far proposed performance evaluation models. For this research execution time and cost of maintenance process are defined as performance measures. For the study, the role based approach is used to understand the flow of Software Maintenance activities in a software organization. Activity diagram is constructed to be transformed into the GSPN. We used PIPE2 to analyze the GSPN. PIPE2 calculates average number of tokens on a place in a GSPN, throughput of timed transition and State Space Analysis. State Space involves calculation of the reachability of the GSPN net using Minimal Siphons and Minimal Traps. This shows whether a GSPN holds the property of Safeness, Boundness and is deadlock free so that a net can be used safely for performance evaluation. After performing the GSPN analysis of Maintenance process results are presented.