Joe McDonnell has posted comments on this change. ( 
http://gerrit.cloudera.org:8080/24159 )

Change subject: IMPALA-14863: Fixes Potential OOM During Row Batch Processing
......................................................................


Patch Set 14:

(1 comment)

http://gerrit.cloudera.org:8080/#/c/24159/14/be/src/exec/partitioned-hash-join-node-ir.cc
File be/src/exec/partitioned-hash-join-node-ir.cc:

http://gerrit.cloudera.org:8080/#/c/24159/14/be/src/exec/partitioned-hash-join-node-ir.cc@60
PS14, Line 60:     ClearExprResultsPool();
> Good insight.  I implemented the solution where the call to ClearExprResult
This is basically the right location. The build side can be arbitrarily large, 
so we can't really move it out to once per probe row. The performance results 
show that the per-row check is below our noise threshold for TPC-H/TPC-DS. So, 
maybe we're done with the old version.

The only thought is that maybe codegen is removing this for most TPC-H/TPC-DS 
queries without any explicit removal from us. So, the question is whether other 
queries could be impacted.

I don't actually know, but maybe the answer is no. If this condition is always 
false, speculative execution should look past it. If this condition is 
happening, other things are more expensive and this is probably not a big deal. 
Maybe simpler is better here.



--
To view, visit http://gerrit.cloudera.org:8080/24159
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-MessageType: comment
Gerrit-Change-Id: Ic54b5c39e1388681275681f22e61b27728dba5af
Gerrit-Change-Number: 24159
Gerrit-PatchSet: 14
Gerrit-Owner: Jason Fehr <[email protected]>
Gerrit-Reviewer: Impala Public Jenkins <[email protected]>
Gerrit-Reviewer: Jason Fehr <[email protected]>
Gerrit-Reviewer: Joe McDonnell <[email protected]>
Gerrit-Comment-Date: Thu, 28 May 2026 17:22:50 +0000
Gerrit-HasComments: Yes

Reply via email to