Is there a way to truncate a graph using Java?
Got this from Jonathan:
https://docs.datastax.com/en/developer/java-driver-dse/1.8/manual/graph/querying/graph_options/
You can run system queries by building a SimpleGraphStatement with the query as a string:
GraphStatement s = new SimpleGraphStatement("system.graph('demo').ifExists().truncate()")
.setSystemQuery();
dseSession.executeGraph(s);
Ah ha! So they have an unpublished truncate method. I'm going to have to try that out.
I wonder what else is hidden on `system.graph()`.
Thanks for sharing.
truncate is a feature of the labs preview version. You won't find it in 6.7 or earlier.
It looks like I can get this effect doing g.V().drop(). Does truncate do more?
it turns out using g.V().drop() with a large enough graph cannot complete. the max-mutations may be exceeded. Using .with("max-mutations", N) can only go as high as the largest integer which may not be large enough to adequately drop all the vertices in a large graph.
Sean, that's what I do. I've got it coded a few different ways.
In my integration tests class, I first make sure that allow_scan is enabled and then run g.V().drop().iterate().
I have a data loader code that uses DSE Analytics with the Java API (and let me tell you, coding Spark with Java is a pain, but I'm in a Java-only shop and Scala wouldn't fly here.). It too uses g.V().drop().iterate(), but it seems to take a long time even with graphs that are just a few GB of data.
I suspect that there's some bookkeeping that DSE Graph is wanting to do in the background which prevent it from just doing a TRUNCATE TABLE operation on each table.
5 People are following this question.
