Automated documentation of Business Domain assignments and cloud application information from an application development pipeline

Abstract

Enterprise Architecture Management (EAM) has emerged to be a settled instrument to increase data quality, reduce IT costs and to reduce the error-prone effects of the process of the Enterprise Architecture information collection.

Although EAM has gained importance over more than a decade to improve the alignment of business with IT and transmit a holistic view of the entire organization along with its application landscape, this discipline has been shown to be more complex than expected.

One of the major challenges for an organization is the documentation of Enterprise Architecture (EA) information. Most of the EA documentation is still collected manually. The high number of applications within the application landscape coupled with data redundancy and the inconsistent data leads to high complexity of EA documentation. This produces a time consuming and highly error-rated documentation and collection process of EA information. Maintaining and gathering information for the EA is (as well) a very expensive task.  The lack of governance is also a major challenge for EA itself. The absence of name conventions, standardized tools integration, continuous delivery and build pipelines, etc., adds to the complexity of a clear and transparent EA documentation.

Since one of the goals of EAM is to create a completely holistic view of the EA, it should integrate external sources, such as cloud services and other repositories.

This thesis will focus on the automated documentation of cloud applications information from an application development pipeline, and the business domain assignments based on the hypothesis that the automated documentation of EA cloud applications leads to a reduction of IT costs and effort, while increasing the data quality of EA information and data.

Once a holistic view is reached, the EA Tool can target a continuous process of monitoring application performance and infrastructure with the already available documented EA information.

To support this, the following research questions will be answered during this thesis.

Research questions:

1. How to obtain EA relevant information from the runtime behavior of cloud-based environments?

2. How to assign the application landscape to business domains?

3. How to automate the assignment process with an integrated toolchain?

4. How does a prototype implementation of the automated documentation process of cloud applications look like?