toggle quoted messageShow quoted text
Answering my own question, BulkLoaderVertexProgram uses a nested instance of a graph in each worker to do the trick.
On Tuesday, May 9, 2017 at 5:36:57 PM UTC+9, Marcelo Freitas wrote:
ok, I got it now that the support GremlinHadoop will have to cassandra is as a reader.
So, what is the recommended way to persist this data back to the cassandra/elasticsearch cluster?
On Tuesday, May 9, 2017 at 3:40:54 PM UTC+9, Marcelo Freitas wrote:
This might be a stupid question, but when running tasks in Hadoop do we have access to the index backend?
If no, what happens when we run something like the PageRankVertexProgram and the indexes we update are actually MixedIndexes? Only Cassandra gets updated? Does the update on elasticsearch side gets trigger automatically somehow?
If yes, do you have any pointers to docs on how to setup this?