Change Support in Adaptive Software: A Case Study for Fine-Grained Adaptation

Adaptive software is a closed-loop system which aims at adjusting itself in different situations at runtime. This paper looks at adaptation as changes in the context of dynamic software evolution, and proposes a conceptual model for these changes based on Activity Theory. This model consists of a hierarchy of activities making changes, and the objectives motivating these changes. This model is an attempt towards establishing a formal framework for designing adaptive software systems. While the proposed model is applicable to any type of adaptation, at different levels of granularity of various software systems, the paper focuses only on fine-grained adaptation changes. As a casestudy, a mission-critical e-commerce system, TPC-W, isused to apply the proposed model and evaluate the effectiveness of fine-grained adaptation changes. The conducted set of experiments aims at evaluating self-optimizing and self-configuring adaptation activities performed through several fine-grained actions such as service-level upgrading/degrading.