February 16, 2015
Graphs is by no question an important step in regard to structure large datasets in more meaningful ways.
I figured that many won’t have the necessary cluster to get the Rexster API up and running, and is most interested in giving graph tech a try.
On the other side Neo4j just isn’t it when planning a large cluster setup with Rexster/Titan. In order to test the API, a local graph may be created with the following Rexster graph configuration (you set this in
<graph> <graph-enabled>true</graph-enabled> <graph-name>testgraph</graph-name> <graph-type>testgraph</graph-type> <graph-location>/tmp/graph.test</graph-location> <extensions> <allows> <allow>tp:gremlin</allow> </allows> </extensions>
Rexster is still executed with
bin/rexster.sh --start. I’ve given a more advanced query example in my code sample mtgx2rexster script migrating all commonly used attributes of Maltego graphml-files. A bit simplified, just to show the concept may be (
pip install rexpro:
from rexpro import RexProConnection graph = RexProConnection('localhost',8184,'testgraph') graph.execute("""g.addVertex(['name':'test1']); g.addVertex(['name':'test2']);""")