Graph does not support the provided graph computer: SparkGraphComputer

Russell Jurney

unread,

Feb 10, 2018, 12:26:05 AM2/10/18

to gremli...@googlegroups.com

I had this code working I thought, but now when I initialize my graph, I get this exception: Graph does not support the provided graph computer: SparkGraphComputer. What do I do? I can't recall how I had this working :/

Is it even possible to load records from Cassandra into JanusGraph in Gremlin and run a Spark operation on them? I swear I did this before, but I can't figure out how I did it now.

The code is:

:plugin use tinkerpop.hadoop
:plugin use tinkerpop.spark

// I edited the keyspace in this file to github_graph
olap_graph = JanusGraphFactory.build().
set("storage.backend", "cassandra").
set("storage.hostname", "127.0.0.1").
set("storage.cassandra.keyspace", "github_graph").
open()

// Get a graph traverser
olap_g = olap_graph.traversal().withComputer(SparkGraphComputer)

// Test things out with a vertex count
// assert(olap_g.V().count().next() == 8139595)

// Add co-forked edges between nodes
edgePairs = olap_g.V().\
hasLabel('repo').\
as('repo1').\
in('forked').\
out('forked').\
where(without('x')).\
as('repo2')