Oh I can reproduce that now. Yup, that's a problem, right there On Thursday, July 16, 2015 at 4:45:58 PM UTC+2, Clark Richey wrote: > > So when I run my tests without the executor I get an error that the > transaction has already completed when I try to examine properties of > created / updated nodes. Creating a new transaction doesn't help. > > Sent from my iPhone > > On Jul 16, 2015, at 04:57, Mattias Persson <mat...@neotechnology.com > <javascript:>> wrote: > > Hi Clark, > > I've converted your groovy code into java and run that in a unit test. As > far as I can see all events trigger as they should. > > What I'm worried about is the pattern that seems to be promoted here, > namely to queue the event processing in an executor service. The > transaction state of a transaction is local to the thread executing the > transaction and simply handing over that state to another thread without > any memory barriers isn't safe, i.e. it may yield unpredictable results. > > My advice to you is to simply scrap the executor service and execute > whatever logic you need within the same thread that gets the > beforeCommit/afterCommit calls. I'll also talk to Max about this, and > update the javadocs. > > Please try this and report back with results. > > Best, > Mattias > > -- > You received this message because you are subscribed to the Google Groups > "Neo4j" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to neo4j+un...@googlegroups.com <javascript:>. > For more options, visit https://groups.google.com/d/optout. > >
-- You received this message because you are subscribed to the Google Groups "Neo4j" group. To unsubscribe from this group and stop receiving emails from it, send an email to neo4j+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
