Master Thesis Christopher Janietz

Enhancing enterprise architecture models using application performance monitoring data.

There has been a continuous decrease of time to market in the IT-industry. Being triggered by external pressure as well as new enabling practices such as DevOps and Microservices, architecture designs suffer more than ever from the problem of outdating. Agile practices support an increasing deviation from architecture designs due to changing business goals or technical implementation difficulties. This research tries to solve these issues by augmenting accessible enterprise architecture documentation with information from runtime instrumentation systems, such as present in modern day service monitoring systems. For this purpose, a model is established which allows to map low level monitoring information from service communication to high level architecture descriptions, such as business domains and applications. By automating the process of identifying differences and visualizing this information appropriately, it shall be possible to draw conclusions such as deviations in the implementation. Furthermore, this research tries to identify the correct part in the architecture design and software development process to enable early reaction as well as correction to such.