You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
At the moment, it seems that only for the g.Triples driver call, in data_access.go, the LIMIT information (stmLimit variable there) is being passed to the driver through the MaxElements field of LookupOptions.
For the other driver calls this MaxElements field of LookupOptions does not seem to be used, and the LIMIT operation is apparently executed solely in the planner level (after all the data was already returned from the driver), truncating the result table with the queryPlan.limit() method. This is not optimal as there was no need to return all that data from the driver just for it to be truncated later, with only a portion of it being useful.
Then, the performance could be a little better if this LIMIT processing was done directly in the driver level, as it already seems to be the case for g.Triples.
N.B. Note that there is a difference between rows in the table result and triples returned from a driver call (for more on that, refer to addTriples in data_access.go). Be attentive when truncating one or the other as for the final result to still be what is expected.
The text was updated successfully, but these errors were encountered:
At the moment, it seems that only for the
g.Triples
driver call, indata_access.go
, theLIMIT
information (stmLimit
variable there) is being passed to the driver through theMaxElements
field ofLookupOptions
.For the other driver calls this
MaxElements
field ofLookupOptions
does not seem to be used, and theLIMIT
operation is apparently executed solely in the planner level (after all the data was already returned from the driver), truncating the result table with thequeryPlan.limit()
method. This is not optimal as there was no need to return all that data from the driver just for it to be truncated later, with only a portion of it being useful.Then, the performance could be a little better if this
LIMIT
processing was done directly in the driver level, as it already seems to be the case forg.Triples
.N.B. Note that there is a difference between rows in the table result and triples returned from a driver call (for more on that, refer to
addTriples
indata_access.go
). Be attentive when truncating one or the other as for the final result to still be what is expected.The text was updated successfully, but these errors were encountered: