Zoltan Borok-Nagy has posted comments on this change. ( 
http://gerrit.cloudera.org:8080/24464 )

Change subject: IMPALA-15076: Fix JDBC scan duplication in UNION ALL and 
ResultSet race
......................................................................


Patch Set 1:

(4 comments)

The change looks good, only have some minor comments.

http://gerrit.cloudera.org:8080/#/c/24464/1//COMMIT_MSG
Commit Message:

http://gerrit.cloudera.org:8080/#/c/24464/1//COMMIT_MSG@28
PS1, Line 28: Testing:
Is it possible to add a planner test for this?


http://gerrit.cloudera.org:8080/#/c/24464/1//COMMIT_MSG@30
PS1, Line 30: impala-private-basic-parameterized
nit: these builds are not known upstream. Instead you could just say you run 
tests in exhaustive mode, using a release build.


http://gerrit.cloudera.org:8080/#/c/24464/1/be/src/scheduling/scheduler.cc
File be/src/scheduling/scheduler.cc:

http://gerrit.cloudera.org:8080/#/c/24464/1/be/src/scheduling/scheduler.cc@608
PS1, Line 608: !fragment_state->scan_range_assignment.empty()
This seems to be a quicker check, so you could check it first.

Same in L878.


http://gerrit.cloudera.org:8080/#/c/24464/1/be/src/scheduling/scheduler.cc@1178
PS1, Line 1178: Must be checked before exec_at_coord
I'm not sure about this. What happens if user uses `set NUM_NODES=1;` to 
enforce non-distributed execution.



--
To view, visit http://gerrit.cloudera.org:8080/24464
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-MessageType: comment
Gerrit-Change-Id: I60ed011faa2177af67ea681c2cd2967648e4a963
Gerrit-Change-Number: 24464
Gerrit-PatchSet: 1
Gerrit-Owner: Arnab Karmakar <[email protected]>
Gerrit-Reviewer: Arnab Karmakar <[email protected]>
Gerrit-Reviewer: Balazs Hevele <[email protected]>
Gerrit-Reviewer: Impala Public Jenkins <[email protected]>
Gerrit-Reviewer: Zoltan Borok-Nagy <[email protected]>
Gerrit-Comment-Date: Wed, 17 Jun 2026 09:16:33 +0000
Gerrit-HasComments: Yes

Reply via email to