Establishing User-centric Cloud Service Registries

Abstract

Many potential cloud consumers are overburdened by the challenges persisting when discovering, assessing, and selecting contemporary Cloud Service offerings: the cloud market is vast and fast-moving, the selection criteria are ambiguous, service knowledge is scattered through the Internet, and features as well as prices are complex and incomparable. Much research has been carried out to create cloud service registries to help users select cloud services for eventual consumption, especially within the field of semantic web services. Through analyzing real-world requirements of six use cases we identified a gap in research for user-centric technologies. We fill this gap by creating a business vocabulary reflecting common service selection criteria, defining a textual domain specific language to let any user describe services easily, and implementing a novel brokering and matchmaking component to support users in their selection process. As a combination of those technologies, we create the Open Service Compendium (OSC), a crowd-sourced cloud service registry. Our evaluation activities highlight how these developments solve real-world challenges in diverse near-production settings. All of this implies that a substantial benefit for service registry users can be created by following a simple architecture that is focused on their concrete needs — instead of aiming for highest sophistication and broadest applicability as observed in many of the related works.