Open iddan opened 4 years ago
Description When executing:
g.V().out(predicate).is(a, b)
It is signifcantly slower than
g.V().has(predicate, a).union(g.V().has(predicate, .b)
I presume it has to do with the way we use indexes for .out().is() which is not optimized like .has()
.out().is()
.has()
I think I know why it behaves that way. It currently only optimizes has with a single value. It can do better by automatically adding a union on quad indexes.
has
Description When executing:
It is signifcantly slower than
I presume it has to do with the way we use indexes for
.out().is()
which is not optimized like.has()