A Model for Transforming Legacy Systems in a client/Server Environment Based on the Unified Modeling Language(UML)

摘要

In this dissertation the researcher developed the methodology for the migration of computer programs from a legacy architecture to client/server architecture. System migrations have failed frequently, and even so-called successful migrations may have serious usability problems. Additional difficulties include missing documentation of the existing program(s), the persons who developed the existing system are not available for consultation, and, frequently, there are important operational and economic issues that must be considered. The client/server environment is quite different from the source environment; the operating system and implementation languages have changed, and system requirements may have been greatly expanded, frequently including the Internet. User interface equipment and techniques are more comprehensive, system response times may be more demanding, significant software system components may be purchased instead of developed in-house, and other elements of the operating theater may be either entirely new or greatly revised. The methodology for developing systems has evolved significantly. In order to make use of the advantages of client/server equipment, new concepts will need to be embodied in the migrated program, such as the use of middleware, object technology to permit the development of higher quality software, and the separation of functionality into server-side and client-side procedures. This dissertation identifies those factors that most critically affect the possibility of success or of failure in the migration. These factors will make it possible to lessen or eliminate the potential for failure. In addition, this dissertation will provide a model for the conversion of legacy systems to more reliable and scalable client/server systems. For this dissertation, the researcher gathered published material relating to the migration of computer systems from one hardware/software platform to a second. Some of the material discussed the conversion process itself. Other material described successes, failures, general techniques and approaches to the migration. Still others discussed nontechnical aspects, including the creation of migration teams and user training. From this material, the most pertinent factors were identified, and from them, a plan of success was developed. That plan of success is this dissertation.