Automated Enterprise Service Bus Based Enterprise Architecture Documentation

Abstract: As a consequence of the increasing requirements on enterprises caused by globalization, fusion of business and IT, new/changing technologies and especially new regulations Enterprise Architecture Management has gained increasing public attention. In this context Enterprise Architecture (EA) can be considered as a holistic view of an enterprise acting as a "collaboration force" between business aspects such as goals, visions, strategies and governance principles as well as IT aspects. For EAs to be useful and to provide business value their formal and comprehensive documentation (data collection and visualization) is essential. Due to the increasing information amount caused by the extending scope on the business as well as on the IT side the EA modelling is cost-intensive and time-consuming. Most of all the documentation is impaired by increasing linking and integration as well as by the striving for automated cross-company business processes. Consequently, the relevant data are widely scattered. Current approaches are largely determined by manual processes which are able to deal only to a limited extent with the new trends. However, a lot of the needed information, above all about the application landscape and the interrelations, is already present in existing Enterprise Service Bus systems used to facilitate the collaboration between applications both within and beyond boundaries. This allows avoiding the expensive task of data collection. This thesis is concerned with the automated documentation of Enterprise Architectures and the support of decision-making using the information provided by an existing Enterprise Service Bus, whereby as a concrete system SAP NetWeaver Process Integration is used. Therefore a tool-aided automated process for the creation and visualization of an Enterprise Architecture model instance is proposed. An essential aspect in this connection is the coverage of the EA information demand set up by EA frameworks such as CySeMoL and ArchiMate, the analysis of the quality as well as the implementation of the logic to transform SAP PI data into EA information. Moreover, several viewpoints are proposed to visualize the extracted data appropriately especially regarding decision problems in the area of the collaboration between applications both within and beyond boundaries.