Migration of Legacy Information Systems

Jean-Luc Hainaut, Anthony Cleve, Jean Henrard, Jean-Marc Hick

Research output: Contribution in Book/Catalog/Report/Conference proceedingChapter

Abstract

This chapter addresses the problem of platform migration of large business applications, that is, complex software systems built around a database and comprising thousands of programs. More specifically, it studies the substitution of a modern data management technology for a legacy one. Platform migration raises two major issues. The first one is the conversion of the database to a new data management paradigm. Recent results have shown that automated lossless database migration can be achieved, both at the schema and data levels. The second problem concerns the adaptation of the application programs to the migrated database schema and to the target data management system. This chapter first poses the problem and describes the State of the Art in information system migration. Then, it develops a two-dimensional reference framework that identifies six representative migration strategies. The latter are further analyzed in order to identify methodological requirements. In particular, it appears that transformational techniques are particularly suited to drive the whole migration process. We describe the database migration process, which is a variant of database reengineering. Then, the problem of program conversion is studied. Some migration strategies appear to minimize the program understanding effort, and therefore are sound candidates to develop practical methodologies. Finally, the chapter describes a tool that supports such methodologies and discusses some real-size case studies.
Original languageEnglish
Title of host publicationSoftware Evolution
EditorsMens Tom, Demeyer Serge
Place of PublicationBerlin Heidelberg
PublisherSpringer
Pages105-138
Number of pages34
ISBN (Print)978-3-540-76439-7
Publication statusPublished - 2008

Fingerprint

Dive into the research topics of 'Migration of Legacy Information Systems'. Together they form a unique fingerprint.
  • Evolution: Evolution

    Cleve, A.

    1/01/1031/01/10

    Project: Research Axis

Cite this