Another thing you should know with the native sparql engine is that it requires a well-populated MetaStructure to work correctly, i.e. a meta model which knows how your data structure looks like (which predicates refers to literals and which refers to other resources). We'll look into how to deal with that dependency. What I'm saying is that you may have problems getting answers with it if you don't supply it.
Did you see my other reply on that I could get results using your first approach? Best, Mattias 2009/2/19 Emil Eifrem <e...@neotechnology.com>: > On Thu, Feb 19, 2009 at 00:39, Lyudmila L. Balakireva <lu...@lanl.gov> wrote: >> But I worry how it going to recover with billions nodes in it . It should >> be able to recover with the same memory used to write the indexes? > > Johan will have to chime in here with the specifics tomorrow (he's > asleep right now). But I believe the memory required here is depending > on the size of whatever non-completed transactions that existed when > the JVM crashed, not the total size of the node store. > > Cheers, > > -- > Emil Eifrém, CEO [e...@neotechnology.com] > Neo Technology, www.neotechnology.com > Cell: +46 733 462 271 | US: 206 403 8808 > _______________________________________________ > Neo mailing list > User@lists.neo4j.org > https://lists.neo4j.org/mailman/listinfo/user > -- Mattias Persson, [matt...@neotechnology.com] Neo Technology, www.neotechnology.com _______________________________________________ Neo mailing list User@lists.neo4j.org https://lists.neo4j.org/mailman/listinfo/user