Github user aray commented on the issue:
https://github.com/apache/spark/pull/18697
Plan for the example query before the patch (with partitioning as suffix):
```
*HashAggregate(keys=[parent#228], functions=[], output=[level2#274])
hashpartitioning(parent#228, 5)
+- *HashAggregate(keys=[parent#228], functions=[], output=[parent#228])
hashpartitioning(parent#228, 5)
+- *Project [parent#228] hashpartitioning(parent#228, 5)
+- *BroadcastHashJoin [level1#270], [son#227], Inner, BuildRight
hashpartitioning(parent#228, 5)
:- *HashAggregate(keys=[parent#228], functions=[],
output=[level1#270]) hashpartitioning(parent#228, 5)
: +- Exchange hashpartitioning(parent#228, 5)
hashpartitioning(parent#228, 5)
: +- *HashAggregate(keys=[parent#228], functions=[],
output=[parent#228]) UnknownPartitioning(0)
: +- *Project [parent#228] UnknownPartitioning(0)
: +- *BroadcastHashJoin [id#266], [son#227], Inner,
BuildRight UnknownPartitioning(0)
: :- *Project [_1#264 AS id#266]
UnknownPartitioning(0)
: : +- *Filter isnotnull(_1#264)
UnknownPartitioning(0)
: : +- *SerializeFromObject [staticinvoke(class
org.apache.spark.unsafe.types.UTF8String, StringType, fromString,
assertnotnull(input[0, scala.Tuple1, true])._1, true) AS _1#264]
UnknownPartitioning(0)
: : +- Scan ExternalRDDScan[obj#263]
UnknownPartitioning(0)
: +- BroadcastExchange
HashedRelationBroadcastMode(List(input[0, string, false]))
BroadcastPartitioning(HashedRelationBroadcastMode(List(input[0, string,
false])))
: +- *Filter (isnotnull(son#227) &&
isnotnull(parent#228)) UnknownPartitioning(0)
: +- InMemoryTableScan [son#227, parent#228],
[isnotnull(son#227), isnotnull(parent#228)] UnknownPartitioning(0)
: +- InMemoryRelation [son#227,
parent#228], true, 10000, StorageLevel(disk, memory, deserialized, 1 replicas),
Statistics(sizeInBytes=148.0 B, hints=none)
: +- *Project [_1#224 AS son#227,
_2#225 AS parent#228] UnknownPartitioning(0)
: +- *SerializeFromObject
[staticinvoke(class org.apache.spark.unsafe.types.UTF8String, StringType,
fromString, assertnotnull(input[0, scala.Tuple2, true])._1, true) AS _1#224,
staticinvoke(class org.apache.spark.unsafe.types.UTF8String, StringType,
fromString, assertnotnull(input[0, scala.Tuple2, true])._2, true) AS _2#225]
UnknownPartitioning(0)
: +- Scan
ExternalRDDScan[obj#223] UnknownPartitioning(0)
+- BroadcastExchange HashedRelationBroadcastMode(List(input[0,
string, false]))
BroadcastPartitioning(HashedRelationBroadcastMode(List(input[0, string,
false])))
+- *Filter isnotnull(son#227) UnknownPartitioning(0)
+- InMemoryTableScan [son#227, parent#228],
[isnotnull(son#227)] UnknownPartitioning(0)
+- InMemoryRelation [son#227, parent#228], true,
10000, StorageLevel(disk, memory, deserialized, 1 replicas),
Statistics(sizeInBytes=148.0 B, hints=none)
+- *Project [_1#224 AS son#227, _2#225 AS
parent#228] UnknownPartitioning(0)
+- *SerializeFromObject [staticinvoke(class
org.apache.spark.unsafe.types.UTF8String, StringType, fromString,
assertnotnull(input[0, scala.Tuple2, true])._1, true) AS _1#224,
staticinvoke(class org.apache.spark.unsafe.types.UTF8String, StringType,
fromString, assertnotnull(input[0, scala.Tuple2, true])._2, true) AS _2#225]
UnknownPartitioning(0)
+- Scan ExternalRDDScan[obj#223]
UnknownPartitioning(0)
```
and after the patch:
```
*HashAggregate(keys=[parent#228], functions=[], output=[level2#274])
UnknownPartitioning(5)
+- Exchange hashpartitioning(parent#228, 5) hashpartitioning(parent#228, 5)
+- *HashAggregate(keys=[parent#228], functions=[], output=[parent#228])
UnknownPartitioning(5)
+- *Project [parent#228] UnknownPartitioning(5)
+- *BroadcastHashJoin [level1#270], [son#227], Inner, BuildRight
UnknownPartitioning(5)
:- *HashAggregate(keys=[parent#228], functions=[],
output=[level1#270]) UnknownPartitioning(5)
: +- Exchange hashpartitioning(parent#228, 5)
hashpartitioning(parent#228, 5)
: +- *HashAggregate(keys=[parent#228], functions=[],
output=[parent#228]) UnknownPartitioning(0)
: +- *Project [parent#228] UnknownPartitioning(0)
: +- *BroadcastHashJoin [id#266], [son#227], Inner,
BuildRight UnknownPartitioning(0)
: :- *Project [_1#264 AS id#266]
UnknownPartitioning(0)
: : +- *Filter isnotnull(_1#264)
UnknownPartitioning(0)
: : +- *SerializeFromObject
[staticinvoke(class org.apache.spark.unsafe.types.UTF8String, StringType,
fromString, assertnotnull(input[0, scala.Tuple1, true])._1, true) AS _1#264]
UnknownPartitioning(0)
: : +- Scan ExternalRDDScan[obj#263]
UnknownPartitioning(0)
: +- BroadcastExchange
HashedRelationBroadcastMode(List(input[0, string, false]))
BroadcastPartitioning(HashedRelationBroadcastMode(List(input[0, string,
false])))
: +- *Filter (isnotnull(son#227) &&
isnotnull(parent#228)) UnknownPartitioning(0)
: +- InMemoryTableScan [son#227,
parent#228], [isnotnull(son#227), isnotnull(parent#228)] UnknownPartitioning(0)
: +- InMemoryRelation [son#227,
parent#228], true, 10000, StorageLevel(disk, memory, deserialized, 1 replicas),
Statistics(sizeInBytes=148.0 B, hints=none)
: +- *Project [_1#224 AS
son#227, _2#225 AS parent#228] UnknownPartitioning(0)
: +- *SerializeFromObject
[staticinvoke(class org.apache.spark.unsafe.types.UTF8String, StringType,
fromString, assertnotnull(input[0, scala.Tuple2, true])._1, true) AS _1#224,
staticinvoke(class org.apache.spark.unsafe.types.UTF8String, StringType,
fromString, assertnotnull(input[0, scala.Tuple2, true])._2, true) AS _2#225]
UnknownPartitioning(0)
: +- Scan
ExternalRDDScan[obj#223] UnknownPartitioning(0)
+- BroadcastExchange HashedRelationBroadcastMode(List(input[0,
string, false]))
BroadcastPartitioning(HashedRelationBroadcastMode(List(input[0, string,
false])))
+- *Filter isnotnull(son#227) UnknownPartitioning(0)
+- InMemoryTableScan [son#227, parent#228],
[isnotnull(son#227)] UnknownPartitioning(0)
+- InMemoryRelation [son#227, parent#228], true,
10000, StorageLevel(disk, memory, deserialized, 1 replicas),
Statistics(sizeInBytes=148.0 B, hints=none)
+- *Project [_1#224 AS son#227, _2#225 AS
parent#228] UnknownPartitioning(0)
+- *SerializeFromObject
[staticinvoke(class org.apache.spark.unsafe.types.UTF8String, StringType,
fromString, assertnotnull(input[0, scala.Tuple2, true])._1, true) AS _1#224,
staticinvoke(class org.apache.spark.unsafe.types.UTF8String, StringType,
fromString, assertnotnull(input[0, scala.Tuple2, true])._2, true) AS _2#225]
UnknownPartitioning(0)
+- Scan ExternalRDDScan[obj#223]
UnknownPartitioning(0)
```
Note there is now an `Exchange` inserted between the two top level
`HashAggregate`
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
