Heterogeneous cloud systems monitoring using semantic and linked data technologies

Abstract

Cloud businesses need comprehensive visibility on hardware and software components, their utilization and their configuration. In addition, they need to integrate such information with their asset management systems and publicly available information such as hardware specifications. In this paper, we present an approach for cloud management and monitoring based on a semantic layer that unifies different interfaces and representations, and makes all relevant information accessible from a single point. We show a proof-of-concept based on OpenStack and Linked Data technologies and evaluate it in terms of overhead, query execution times, and effectiveness in the data gathering phase. Our findings indicate that a semantics-based approach is indeed feasible and advantageous for providing uniform access across different cloud environments and levels.