-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/37128/
-----------------------------------------------------------

Review request for hive and Xuefu Zhang.


Bugs: HIVE-11180
    https://issues.apache.org/jira/browse/HIVE-11180


Repository: hive-git


Description
-------

The improvement was introduced in HIVE-9824. Let's use this task to track how 
we can enable that for spark.


Diffs
-----

  itests/src/test/resources/testconfiguration.properties c710b0b 
  
ql/src/java/org/apache/hadoop/hive/ql/exec/persistence/MapJoinTableContainerSerDe.java
 e97a9f0 
  ql/src/java/org/apache/hadoop/hive/ql/exec/spark/HashTableLoader.java 10e3497 
  
ql/src/java/org/apache/hadoop/hive/ql/exec/vector/mapjoin/VectorMapJoinCommonOperator.java
 87ebcf2 
  
ql/src/java/org/apache/hadoop/hive/ql/exec/vector/mapjoin/fast/VectorMapJoinFastTableContainer.java
 f2080f4 
  ql/src/java/org/apache/hadoop/hive/ql/optimizer/physical/Vectorizer.java 
82c3e50 
  
ql/src/java/org/apache/hadoop/hive/ql/optimizer/spark/SparkMapJoinOptimizer.java
 39d1f18 
  ql/src/test/results/clientpositive/spark/vector_inner_join.q.out PRE-CREATION 
  ql/src/test/results/clientpositive/spark/vector_outer_join0.q.out 
PRE-CREATION 
  ql/src/test/results/clientpositive/spark/vector_outer_join1.q.out 
PRE-CREATION 
  ql/src/test/results/clientpositive/spark/vector_outer_join2.q.out 
PRE-CREATION 
  ql/src/test/results/clientpositive/spark/vector_outer_join3.q.out 
PRE-CREATION 
  ql/src/test/results/clientpositive/spark/vector_outer_join4.q.out 
PRE-CREATION 
  ql/src/test/results/clientpositive/spark/vector_outer_join5.q.out 
PRE-CREATION 

Diff: https://reviews.apache.org/r/37128/diff/


Testing
-------

Verified that the newly added test golden files have same results as the Tez 
version. Since hive.vectorized.execution.mapjoin.native.enabled is on by 
default, other cases (map join involving Orc tables and vectorization enabled) 
should be covered automatically.

On the other hand, some tests that may involve the optimization haven't been 
enabled for spark, e.g. vector_char_mapjoin1.q. We can consider enable these 
tests in follow on task.


Thanks,

Rui Li

Reply via email to