CODEWITHSUNDEEP
hadoopneo4j
user1984382
user1984382

Reputation: 33

How to use neo4j as input to hadoop?

I have a large neo4j database. I need to check for multiple patterns existing across the graph, which I was thinking would be easily done in hadoop. However, I'm not sure of the best way to feed tuples from neo4j into hadoop. Any suggestions?

Upvotes: 0

Views: 633

Answers (1)

Vidya
Vidya

Reputation: 30320

In my opinion, while it can be done, I don't think MapReduce (which I believe is what you mean when you say "Hadoop") is a good (or at least performant) choice for graph analytics. You want a Bulk Synchronous Parallel approach instead. If you want to perform cloud-scale graph analytics, you want Apache Giraph, which "understands" the Hadoop ecosystem.

Then again, I would ask why you need to use anything outside of Neo4J at all. I don't know your use case obviously, but first make sure you can't do what you need to do within Neo4J.

Upvotes: 3

Related Questions