impala-reviews mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Dan Hecht (Code Review)" <>
Subject [Impala-ASF-CR] IMPALA-5036: Parquet count star optimization
Date Wed, 21 Jun 2017 20:48:40 GMT
Dan Hecht has posted comments on this change.

Change subject: IMPALA-5036: Parquet count star optimization

Patch Set 5:

File fe/src/main/java/org/apache/impala/catalog/

Line 886:         prefix + "8InitZeroIN10impala_udf9BigIntValEEEvPNS2_15FunctionContextEPT_",
rather than defining a new builtin, why don't we wrap the sum() with a coalesce or isnull?
does that not work for some reason?
File fe/src/main/java/org/apache/impala/planner/

PS5, Line 105:  The caller passes in a flag to the constructor that indicates if the count(*)
             :  * optimization can be applied to the query block of this scan. This scan node
             :  * whether to apply the optimization or not
This doesn't make a lot of sense to me. On one hand, it sayst he caller decides, but then
it says that this node also decides. What's actually going on?

Line 109:  *
do we actually explain in some comment how this optimization works at a high level?

PS5, Line 140:  This
             :   // scan does additional analysis in init() to determine whether it is correct
to apply
             :   // the optimization.
this seems to contradict the comment above that says the caller passes a flag to the constructor
that indicates whether the optimization can be applied.

To view, visit
To unsubscribe, visit

Gerrit-MessageType: comment
Gerrit-Change-Id: I536b85c014821296aed68a0c68faadae96005e62
Gerrit-PatchSet: 5
Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-Owner: Taras Bobrovytsky <>
Gerrit-Reviewer: Alex Behm <>
Gerrit-Reviewer: Dan Hecht <>
Gerrit-Reviewer: Lars Volker <>
Gerrit-Reviewer: Marcel Kornacker <>
Gerrit-Reviewer: Mostafa Mokhtar <>
Gerrit-Reviewer: Taras Bobrovytsky <>
Gerrit-Reviewer: Zach Amsden <>
Gerrit-HasComments: Yes

View raw message