I have a set of document vertices and metadata vertices. Each document is
connected to many metadatas, and each metadata is connected to many
documents.
Each edge has a weight associated with it
I am having trouble with a scalable way to count the vertex connectivity of
the metadata vertices
I've essentially got something like
SELECT sum(weight), name (SELECT weight, outV().name as name, outV() as
metadata (SELECT expand(inE("metadata_of")) FROM #11:0)) ORDER BY name
This returns in ~2 seconds
SELECT sum(weight), name, metadata.inE().size() (SELECT weight, outV().name
as name, outV() as metadata (SELECT expand(inE("metadata_of")) FROM #11:0))
ORDER BY name
I have not yet had this successfully return (waits of 20+ minutes before I
kill it).
f I pull up a metadata object like
SELECT FROM #17:0 (this is the first metadata edge)
I get an object back in the web UI with properties like this under
in.metadata_of:
#18:0 #18:22701 #18:23053 #18:35613 #18:54456 #18:57228 ..More(12054)
And a runtime of 0.383 sec.
When I do something like:
SELECT inV().name FROM (select expand(outE("metadata_of")) from #11:1)
I get 17 results back in 0.744 seconds
When I do:
select inE().size() from #17:0
I get one result back in between 0.1 and 1.0 seconds, most landing around
0.8 seconds.
When I do
SELECT inV().inE().size() FROM (select expand(outE("metadata_of")) from
#11:1)
I have not gotten a return in 120+ seconds.
I'm going to try inserting indexes on the in/out properties for the edges,
but this is slow going as I have a few hundred million edges to deal with.
Should this help? Is there a more sensible way for me to be doing this?
--
---
You received this message because you are subscribed to the Google Groups
"OrientDB" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
For more options, visit https://groups.google.com/d/optout.
