Distributed quality-attribute optimization of software architectures

• Alejandro Rago ISISTAN-UNICEN
• Santiago Vidal ISISTAN-UNICEN
• J. Andres Diaz-Pace ISISTAN-UNICEN
• Sebastian Frank University of Stuttgart
• André van Hoorn University of Stuttgart

Resumo

A key challenge of software architecture design is how to satisfy quality-attribute requirements, which often conflict with each other. This is usually a complex task, because there are several candidates for architectural solutions meeting the same requirements, and quality-attribute tradeoffs of those solutions need to be considered by the architects. In this context, we present the SQuAT framework to assist architects in the exploration of design solutions and their tradeoffs. This framework provides a modular approach for integrating quality-attribute analyzers and solvers, and also features a distributed search-based optimization. In this paper, we report on an experience using SQuAT with Palladio architectural models, which integrates third-party tools for performance and modifiability, and shows the tradeoffs among candidate solutions to the architect. Furthermore, we enhance the standard search schema of SQuAT with a distributed negotiation technique based on monotonic concession, in order to provide better tradeoffs for the architect's decision making.

Palavras-chave: software architectures, quality attributes, agents