ulysses-you commented on a change in pull request #32450:
URL: https://github.com/apache/spark/pull/32450#discussion_r627136151
##########
File path:
sql/catalyst/src/main/scala/org/apache/spark/sql/internal/SQLConf.scala
##########
@@ -598,6 +598,16 @@ object SQLConf {
.bytesConf(ByteUnit.BYTE)
.createOptional
+ val ADAPTIVE_SHUFFLE_HASH_JOIN_LOCAL_MAP_THRESHOLD =
+ buildConf("spark.sql.adaptive.shuffledHashJoinLocalMapThreshold")
+ .doc("Configures the maximum size in bytes for per partition that can be
allowed to build " +
+ "local hash map. If all the partition size not larger than this
threshold, join " +
+ s"selection may use shuffled hash join instead of sort merge join when
" +
+ s"${PREFER_SORTMERGEJOIN.key} is false.")
+ .version("3.2.0")
+ .bytesConf(ByteUnit.BYTE)
+ .createWithDefaultString("64MB")
Review comment:
The main idea is that the default skew join size is `256MB` and the
local map should smaller 3x(follow the existed formula) than other side. So
assume the local map size is `64MB` and other side is `192MB`.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
