Actually there are two options with the batch-importer: #1 faster: you drive the node-id generation externally using :id on a column with dedicated node-ids and then for the relationship use start:id and end:id as first columns pointing to the node-id's from your nodes files
#2 use indexes, slower: configure your nodes to be indexed,e.g. using as header name:string:Users and in the rels-csv files use name:string:Users in both the start and end column (first and second col), then configure the index in your batch.properties and you're good to go On Mon, Mar 24, 2014 at 1:14 PM, natalie Yosef <[email protected]> wrote: > hey, > i'm trying to import csv files to neo4j > i have one csv with 7,000,000 rels > and another with 250,000,000 rels > i tries importing the data with the java api but after 7,000,000 rels it > gets stack on > "unable to get transaction" and also it takes lots of time - about 2 days > > i tried using the batch inserter - the one on github batch_impoter_20 and > also write one of my one, > the problem is I have few rels csvs, so i want to find the entites > allready created and attach them, > but there is no such option in the batchInserter (to see if node allready > exist) and then create a relation between them > > i'm really stack any idea on either approach will be usefull > thank you > > > -- > 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. > -- 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.
