HyukjinKwon commented on a change in pull request #24618:
[SPARK-27734][CORE][SQL] Add memory based thresholds for shuffle spill
URL: https://github.com/apache/spark/pull/24618#discussion_r390246347
##########
File path: core/src/main/scala/org/apache/spark/internal/config/package.scala
##########
@@ -998,6 +998,26 @@ package object config {
.intConf
.createWithDefault(Integer.MAX_VALUE)
+ private[spark] val SHUFFLE_SPILL_MAP_MAX_SIZE_FORCE_SPILL_THRESHOLD =
+ ConfigBuilder("spark.shuffle.spill.map.maxRecordsSizeForSpillThreshold")
+ .internal()
+ .doc("The maximum size in memory before forcing the map-side shuffle
sorter to spill. " +
+ "By default it is Long.MAX_VALUE, which means we never force the
sorter to spill, " +
+ "until we reach some limitations, like the max page size limitation
for the pointer " +
+ "array in the sorter.")
+ .bytesConf(ByteUnit.BYTE)
+ .createWithDefault(Long.MaxValue)
Review comment:
@amanomer, you can make this configuration optional via `createOptional` to
represent no limit.
----------------------------------------------------------------
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]
With regards,
Apache Git Services
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
