Zoltan Borok-Nagy has posted comments on this change. ( 
http://gerrit.cloudera.org:8080/16720 )

Change subject: IMPALA-10325: Parquet scan should use min/max statistics to 
skip pages based on equi-join predicate
......................................................................


Patch Set 50:

(5 comments)

http://gerrit.cloudera.org:8080/#/c/16720/50//COMMIT_MSG
Commit Message:

http://gerrit.cloudera.org:8080/#/c/16720/50//COMMIT_MSG@68
PS50, Line 68: partition columns
It will be especially useful for partitioned Iceberg tables because runtime 
bloom filters doesn't work with them.

But they'll need some special code, because partitioned Iceberg tables are 
treated as non-partitioned tables most of the time.


http://gerrit.cloudera.org:8080/#/c/16720/50/be/src/exec/parquet/hdfs-parquet-scanner.cc
File be/src/exec/parquet/hdfs-parquet-scanner.cc:

http://gerrit.cloudera.org:8080/#/c/16720/50/be/src/exec/parquet/hdfs-parquet-scanner.cc@659
PS50, Line 659: raio
nit: ratio


http://gerrit.cloudera.org:8080/#/c/16720/50/be/src/runtime/runtime-filter-ir.cc
File be/src/runtime/runtime-filter-ir.cc:

http://gerrit.cloudera.org:8080/#/c/16720/50/be/src/runtime/runtime-filter-ir.cc@33
PS50, Line 33: min_max_filter_.Load() == nullptr) return true;
             :       return min_max_filter_.Load()
Probably we should only invoke Load() once and store the pointer.


http://gerrit.cloudera.org:8080/#/c/16720/50/fe/src/main/java/org/apache/impala/planner/HdfsScanNode.java
File fe/src/main/java/org/apache/impala/planner/HdfsScanNode.java:

http://gerrit.cloudera.org:8080/#/c/16720/50/fe/src/main/java/org/apache/impala/planner/HdfsScanNode.java@733
PS50, Line 733: referes
nit: refers


http://gerrit.cloudera.org:8080/#/c/16720/50/fe/src/main/java/org/apache/impala/planner/HdfsScanNode.java@734
PS50, Line 734: as it will not be
              :     // as effective as the conjunct.
How do we know the runtime efficiency here? Maybe we should still add it and 
let minmax threshold decide?
However, if the min/max conjunct comes from an EQ predicate, then yes, we 
should probably not add an overlap predicate in this case.



--
To view, visit http://gerrit.cloudera.org:8080/16720
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-MessageType: comment
Gerrit-Change-Id: I379405ee75b14929df7d6b5d20dabc6f51375691
Gerrit-Change-Number: 16720
Gerrit-PatchSet: 50
Gerrit-Owner: Qifan Chen <qc...@cloudera.com>
Gerrit-Reviewer: Csaba Ringhofer <csringho...@cloudera.com>
Gerrit-Reviewer: Impala Public Jenkins <impala-public-jenk...@cloudera.com>
Gerrit-Reviewer: Qifan Chen <qc...@cloudera.com>
Gerrit-Reviewer: Tim Armstrong <tarmstr...@cloudera.com>
Gerrit-Reviewer: Zoltan Borok-Nagy <borokna...@cloudera.com>
Gerrit-Comment-Date: Thu, 21 Jan 2021 15:55:15 +0000
Gerrit-HasComments: Yes

Reply via email to