Abstract : One of the key elements of the Semantic Web is the Resource Description Framework (RDF). Efficient storage and retrieval of RDF data in large scale settings is still challenging and existing solutions are monolithic and thus not very flexible from a software engineering point of view. In this paper, we propose a modular system, based on the scalable Content-Addressable Network (CAN), which gives the possibility to store and retrieve RDF data in large scale settings. We identified and isolated key components forming such system in our design architecture. We have evaluated our system using the Grid'5000 testbed over 300 peers on 75 machines and the outcome of these micro-benchmarks show interesting results in terms of scalability and concurrent queries.