>> After playing with it more, the culprit seems to be not the symbolic >> link but the local drive :-/ When I delete the neo4j directory with the >> db and start the application from the local drive then I'm getting the >> NPEs - even if I shut the server down and start it again. When I delete >> the neo directory and start jboss from the directory on the nfs drive >> then it works fine, without NPEs. When I then start the application from >> the local drive again (without deleting the neo dir) then it again works >> fine, without NPEs. >> I have no idea why this is happening.. I verified that the index service >> is being initialized in the same thread that uses it later when the >> exception occurs so it shouldn't be a problem with some asynchronicity >> introduced by JBoss. >>
That is quite strange. I've never come across a problem like this and we don't have automated tests for situations like these. Could you perhaps provide me with the stacktrace you get? That would be very helpful. > > One last note: the described behaviour holds for the combination of > index-util-0.9-20091112.150844-2.jar > neo-1.0-b11-20091114.181002-8.jar > > with > index-util-0.9-20091116.160830-4.jar > neo-1.0-b11-20091114.181002-8.jar > > I get NPEs in both cases (local/nfs). Which combination should I > actually be using? The best thing is to use the index-util-0.9-SNAPSHOT and neo-1.0-b11-SNAPSHOT versions (i.e. the latest of both those components). _______________________________________________ Neo mailing list User@lists.neo4j.org https://lists.neo4j.org/mailman/listinfo/user
