Federated or cross graph queries

We have a scenario where we have an RDF file provided by a third party, which we've imported in to Neo4j using the Neosemantics plugin. The third party RDF file is regularly updated so we wanted to host this graph with little to no changes.

However we do want to extend and embellish the third party graph with our own vertices and edges. So, the question is:

    • Can you do cross graph queries (or federated in SPARQL terms) to return values from both graphs?
    • If this is not possible what are the best approaches for extended an existing graph without impacting the original and still allowing for updates/imports?

Any help would be greatly received.


The timing here is pretty good, as we've been making more announcements this month on the new features being introduced in the upcoming Neo4j 4.0 release. This will include multi-graph, multiple simultaneously-running graphs on a single database, and there are some yet to be fully-announced features that speak directly to your needs here.

Wait a little bit longer this week and watch for Neo4j announcements, we have more to talk about regarding these new features.

1 Like

Thanks Andrew, I eagerly await the announcement.

Many thanks for your prompt response.

That’s awesome! Can’t wait multiple graph queries!

Lost track of this one!

If you were paying attention to the NODES conference and Neo4j announcements for 4.0, you likely saw that we're introducing Neo4j Fabric with 4.0, which allows a proxy Fabric database to be configured to query against other databases, which may be separate remote dbms (single or cluster), or databases within a single dbms (this can go hand-in-hand with multi-database functionality). The Fabric proxy db can even be on the same dbms as the multiple databases that it has access to (1 machine, 1 dbms, multiple databases including the Fabric database).

See the linked documentation on how to query across multiple databases with Fabric!