[eranation]

Thanks for the post! You mentioned Spark via OLAP connectivity. Can you please elaborate a little on how gremlin works with spark? Does it use the GraphX API behind the scenes or is it just spark? Are there any sources on how well it works?

[okram]

Gremlin (over Spark) does not use GraphX. It simply represents the graph as a tensor RDD (i.e. a multi-layered matrix) and with the Spark functional library, it implements BSP-based vertex-centric computing (i.e. message passing). You can see examples and a diagram explaining how it works at this location:

http://tinkerpop.incubator.apache.org/docs/3.0.0-incubating/...