Application Of Iterative Process Optimization In Reengineering Of Legacy Systems

It has been a golden rule for companies in nowadays that "Profit maximization, cost minimization." When applying this rule to the field of software engineering, it requires a higher qualified product with lower costs, which means a short development life cycle is compatible. Hence the iterative model is popular in the reengineering process. The fact is lots of rework can decay the iterative reengineering process of a financial legacy system, considering its complex business logic and strong dependency to the requirements.In this paper, the author focuses on the research of iterative process and proposed methods to optimize the reengineering process. The research develops a two-layer method for the iterative reengineering process and is brought into practice in the reengineering of legacy financial system V:(1) The first layer is for the iterative planning of the whole system. By system analysis and business rule extraction, the system can be divided into several modules according to the functionality they act. Then, a continuous DSM, which is different from traditional discrete DSM, is used to analyze the coupling between modules. Cluster Analysis tool is used to classify the modules into different iterations of reengineering process(2) The second layer is to optimize the reengineering sequence of sub modules as well as to assign them to developing groups in each iteration. Rules for task assignment is raised in the paper. Also, an improved algorithm based on the Time-Order based Tearing algorithm is used to produce the optimized developing sequence of sub modules.(3) The two-layer optimization method is applied in the reengineering process of legacy system V. Special characters of financial legacy systems are taken into consideration in the system analysis. The optimized iterative plan of system V is obtained in the reengineering process.Implementing the optimization method can enhance the iterative process of legacy systems. Since product in iteration 1 has been successfully launched, the iterative process planning method is proved to be practical.