Kimahriman commented on a change in pull request #35085:
URL: https://github.com/apache/spark/pull/35085#discussion_r829640386
##########
File path: core/src/main/scala/org/apache/spark/internal/config/package.scala
##########
@@ -654,6 +654,16 @@ package object config {
.booleanConf
.createWithDefault(false)
+ private[spark] val SHUFFLE_SERVICE_REMOVE_SHUFFLE_ENABLED =
+ ConfigBuilder("spark.shuffle.service.removeShuffle")
+ .doc("Whether to use the ExternalShuffleService for deleting shuffle
blocks for " +
+ "deallocated executors when the shuffle is no longer needed. Without
this enabled, " +
+ "shuffle data on executors that are deallocated will remain on disk
until the " +
+ "application ends.")
+ .version("3.3.0")
Review comment:
We had to disable the external shuffle service for basically all of our
large shuffling jobs until I got this working, and based on others interest and
all the jira tickets that have been made over the years I'd consider this a
critically missing feature. I'm fine disabling by default, especially if we try
to limit the permission changes anymore, the more likely this will cause issues
using the external shuffle service
--
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.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
