Spinner

 - SPARQL 1.1 Distributed Query Processor sitting on your network

 - Maintains the location and capabilities of any number of SPARQL end points

 - Virtualizes data in all end points, making them appear a single collection for querying

 - Plans and optimizes SPARQL queries over any collection of SPARQL end points

Spinner is a component of our Distributed Information Management System (DIMS). Using semantics described by any OWL ontology as if all of the endpoints are a single graph, Spinner implements the federation capabilities provided by the SPARQL 1.1 specification to enable querying of any collection of SPARQL endpoints.

A single Spinner Server can host multiple federation services, each exposed through its own endpoint (URL, derived from that of the base server URL through configuration settings).  Each federation service endpoint exposes a REST API for querying the service, and Web UI tools for querying.

Each federation service exposes one or more external SPARQL services, or sources.  Each of these source services is referred to as a federated service in the context of the federation service in which it is configured.  A given source service can be used multiple times in different federation services, configured as different federated services in each context.  The following figure illustrates a Spinner Server hosting multiple federation services, each configured with multiple federated services, with one source service used by both federation services.

The Spinner "Early Availability" (EA) release is now available, and may be downloaded from the Spinner download page.