Domain-Driven Design of Big Data Systems based on a Reference Architecture

In general, different application domains may require different big data systems. To enhance the understanding of big data systems and support the architect in designing big data architectures, we propose a domain-driven design approach for deriving application architectures. To this end, we propose a domain engineering approach in which a family feature model, reference architecture, and corresponding design rules are identified. The family feature model is derived based on a domain analysis of big data systems and represents the common and variant features. The reference architecture represents a generic structure for various application architectures of big data systems. Finally, the design rules define reusable design heuristics for designing an application architecture based on the selection of features of the family feature model and the reference architecture. We illustrate our approach for deriving the big data architectures of different well-known big data systems.