Evaluating Software Architectures for Stability: A Real Options Approach

Architectural stability refers to the extent an architecture is flexible to endure evolutionary changes in stakeholders’ requirements and the environment, while leaving the architecture intact. In an evolutionary context, there is a pressing need for stable software architectures. In this context, requirements are generally volatile; they are likely to change and evolve over time. The change is inevitable as it reflects changes in stakeholders’ needs and the environment in which the software system works. The tension between an unstable architecture and the volatile requirements may entail large and disruptive changes for the requirements to be accommodated. The change may “break” the architecture necessitating changes to the architectural structure (e.g. changes to components and interfaces), architectural topology (e.g. architectural style, where a style is a generic description of a software architecture), or even changes to the underlying architectural infrastructure (e.g. middleware). It may be expensive and difficult to change the architecture as requirements evolve [6]. Consequently, failing to accommodate the change leads ultimately to the degradation of the usefulness and the value of the system. From an economic perspective, the volatility of requirements is a source of uncertainty that places the long-term investment in a particular architecture at risk. If the business goal that the system should be long-lived, should evolve to accommodate future changes, and should create future value, stability becomes an important architectural quality to evaluate an architecture for. The evaluation is necessary to cope with the incomplete knowledge in an evolutionary context and mitigate risks in the long-term investment in a particular architecture. The evaluation is crucial for analyzing trade-offs between two or more candidate software architectures for stability; analyzing the strategic position of the enterpriseif the enterprise is highly centered on the software architecture (as it is the case in web-based service providers companies: e.g. amzon.com); valuing the long-term investment in a particular architecture; and validating the architecture for evolution. Our work addresses the following research question: can we use an economic approach (real-options theory) to systematically evaluate the stability of an architecture in the face of the changing requirements? The abstract is further structured as follows. Section 2 presents our research claims. Section 3 summarizes our approach in exploiting options theory to evaluate architectural stability. Section 4 discusses related work. Section 5 summarizes our work in progress and expected contributions.