Distributed RDF Engine with Adaptive Query
Optimization & Minimal Communication
In this work, we promote a new classification of paradigms by which RDF systems can be designed. In particular, we suggest that RDF systems can be built in four different ways as portrayed in Figure 1.
DREAM adopts a master-slave architecture.Each slave machine can, in principle, encompass any centralized RDF store.
DREAM employs the Quadrant-IV paradigm. As such, it stores an RDF dataset unsliced at each cluster machine and adopts a graph-based, rule-oriented query planner.
We fully implemented DREAM using C and MPICH 3.0.4, and evaluated it on a private and a public (i.e., Amazon EC2) clouds.