Many systems are built with very little foresight for incorporating changes resulting from new requirements. This paper provides a systematic way and normative guidance to analyze the variations in SOA solution design. During SOA solution design, two types of changeoriented analysis patterns, variation-oriented analysis (VOA) and event-driven change analysis (EDCA), are proposed to create a resilient design that is adaptive to changes. Specifically, a meta-data model and associated assets for solution design are introduced first to capture the information context for SOA solution design. Then a new change propagation analysis process is presented to summarize the variation identification, variation impact assessment, variation impact path identification, variation analysis and decision making. Based on these two techniques, this paper creates a set of normative guidance for the variation oriented design. These guidelines include but not limited to the artifact type, variation description, variation type, variation artifacts, and impact analysis.
Citation:
Liang-Jie Zhang, Ali Arsanjani, Abdul Allam, Dingding Lu, Yi-Min Chee, "Variation-Oriented Analysis for SOA Solution Design," scc, pp.560-568, IEEE International Conference on Services Computing (SCC 2007), 2007