Github user hbdeshmukh commented on a diff in the pull request:
https://github.com/apache/incubator-quickstep/pull/41#discussion_r68635258
--- Diff: relational_operators/HashJoinOperator.cpp ---
@@ -65,21 +65,11 @@ DEFINE_bool(vector_based_joined_tuple_collector, false,
"If true, use simple vector-based joined tuple collector in "
"hash join, with a final sort pass to group joined tuple pairs
"
"by inner block. If false, use unordered_map based collector
that "
- "keeps joined pairs grouped by inner block as they are found "
- "(this latter option has exhibited performance/scaling
problems, "
- "particularly in NUMA contexts).");
+ "keeps joined pairs grouped by inner block as they are
found.");
--- End diff --
I don't mind removing it. If I recall correctly, an older profiling done by
@pateljm revealed that map collector outperformed the vector collector pretty
much always.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
