BatchInserter.shutdown() is synchronous, it has to be, otherwise it wouldn't be safe at all. After shutdown() has returned it guarantees that all store files have been written, and are in a consistent state.
-t On Mon, Feb 14, 2011 at 4:01 PM, Pablo Pareja <[email protected]> wrote: > Great, thanks for the information. > Just out of curiosity then, are calls to BatchInserter.shutdown() sync or > asynchronous ? > I mean, does the program flow stops in that call till every operation the > shutdown implies has been > performed? > Cheers, > > Pablo > > On Mon, Feb 14, 2011 at 3:30 PM, Tobias Ivarsson < > [email protected]> wrote: > > > The batch inserter will write data to disk as needed. The memory mapped > > regions are used to speed up writes, so that the API level writes can be > > done to the memory mapped regions which can then be written in large > > sequential chunks to disk. > > > > You do not have to shut down the batch inserter for this to happen, it is > > taken care of automatically. > > > > Cheers, > > Tobias > > > > On Mon, Feb 14, 2011 at 10:53 AM, Pablo Pareja <[email protected]> wrote: > > > > > Hi, > > > > > > I was wondering whether once the memory-mapping configuration for > > > BatchInserter is full *(for example* > > > *in the case of neostore.relationshipstore.db.mapped_memory) *the > class > > > would auto-commit these changes > > > to disk or it would just throw an exception. > > > In the latter case, what would the options be? > > > > > > In my case I have to insert a huge amount of data that cannot be > > allocated > > > in RAM. > > > First I thought the BatchInserter would deal with this kind of > situation, > > > however I keep getting exceptions whenever the > > > program reaches the point where *(I guess) *memory mapping > configuration > > is > > > not enough. > > > > > > Since I cannot use more RAM than what I have, I thought, ok so I will > > > shutdown both inserter and index service and > > > create them again each time the program have already inserted a decent > > > amount of data. > > > Then, don't know why but after a couple of tests, at some point the > > program > > > always halts throwing a really weird exception > > > saying something about a index file that exists but cannot be accessed > in > > > the merge process ?!? > > > > > > I guess I'd be doing something wrong but cannot figure out what it is. > > > Thanks in advance > > > > > > -- > > > Pablo Pareja Tobes > > > LinkedIn http://www.linkedin.com/in/pabloparejatobes > > > Twitter http://www.twitter.com/pablopareja > > > > > > http://www.ohnosequences.com > > > _______________________________________________ > > > Neo4j mailing list > > > [email protected] > > > https://lists.neo4j.org/mailman/listinfo/user > > > > > > > > > > > -- > > Tobias Ivarsson <[email protected]> > > Hacker, Neo Technology > > www.neotechnology.com > > Cellphone: +46 706 534857 > > _______________________________________________ > > Neo4j mailing list > > [email protected] > > https://lists.neo4j.org/mailman/listinfo/user > > > > > > -- > Pablo Pareja Tobes > LinkedIn http://www.linkedin.com/in/pabloparejatobes > Twitter http://www.twitter.com/pablopareja > > http://www.ohnosequences.com > _______________________________________________ > Neo4j mailing list > [email protected] > https://lists.neo4j.org/mailman/listinfo/user > -- Tobias Ivarsson <[email protected]> Hacker, Neo Technology www.neotechnology.com Cellphone: +46 706 534857 _______________________________________________ Neo4j mailing list [email protected] https://lists.neo4j.org/mailman/listinfo/user
