Cloud services composition through cloud patterns: a semantic-based approach

Despite the research and standardization efforts carried out both by academia and commercial enterprises, the composition of existing cloud services which fully satisfy customers' requirements is still a complex and tricky task. This situation is due to the high number of cloud services currently available on the market, which either expose non-standard interfaces or implement different standards, according to their specific objective. As in the past design patterns have been applied to software design to bring order and help developers in better building, composing and reusing their application, nowadays cloud patterns offer the opportunity to leverage best practices in services composition to ease the design and deployment of cloud-oriented applications. However, due to differences in semantics which affect services', operations' and parameters' descriptions, cloud patterns alone cannot solve the cloud service composition problem. In this paper a methodology for the discovery and composition of cloud services, guided by cloud patterns, is presented. Such a methodology is supported by semantic Web technologies, such as OWL, OWL-S and SPARQL, to solve incongruence between interfaces' and parameters' descriptions, and to automatize the whole composition process.