Michael, sorry to answer so late.

After following your instructions I got that performance was acceptable.

Thank you very much for your help.

Regards,
Alberto.

On Monday, March 30, 2015 at 8:13:36 AM UTC+2, Michael Hunger wrote:
>
> That's what I said.
>
> Use an effective cache (i.e. one of the primitive collection libraries 
> with a map from long -> long)
>
> Most memory efficient and performant way:
>
> Alternatively, what you do is to do a dual-pass.
>
> Create an Array of the expected sizes, add the key entries to the array.
> Sort the array
> The keys are entries of the array and the array-index is the node-id.
> you can scan the array for duplicates and null them out.
> And then you can use Arrays.binarySearch() to find your entries.
>
> This is quite efficient and similar to what Neo4j uses internally for 
> neo4j-import.
>
> Michael
>
> Am 29.03.2015 um 18:50 schrieb Alberto Jesús Rubio Sánchez <
> [email protected] <javascript:>>:
>
> Hi Michael,
>
> I've been testing and my problem is that the file is very large and the 
> memory becomes full.
>
> For this reason I thought to use a cache to store the ids. If a node id 
> isn't in the cache, the node is inserted  even if the node is in the 
> database. Finally look for duplicate nodes remaining to merge them.
>
> I think it may be a good solution. What do you think?
>
> Thanks,
> Alberto.
>
> -- 
> 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 [email protected] <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 [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to