Computing and Informatics

In this paper we present a semantic-based approach for supporting data-intensive applications in distributed environments. The approach is characterized by usage of explicit definition of non-functional quality parameters regarding storage systems, semantic descriptions of the available storage infrastructre and monitoring data concering the infrastructure workload and users operation, along with an implementation of the approach in the form of a toolkit called FiVO/QStorMan. In particular, we describe semantic descriptions, which are exploited in the storage resource provisioning process. In addition, the paper describes results of the performed experimental evaluation of the toolkit, which confirm the effectiveness of the proposed approach for the storage resource provisioning.