srowen commented on code in PR #50020:
URL: https://github.com/apache/spark/pull/50020#discussion_r1981419793
##########
core/src/main/scala/org/apache/spark/BarrierCoordinator.scala:
##########
@@ -80,8 +81,9 @@ private[spark] class BarrierCoordinator(
states.forEachValue(1, clearStateConsumer)
states.clear()
listenerBus.removeListener(listener)
- ThreadUtils.shutdown(timer)
} finally {
+ timerFuture.foreach(_.cancel(true))
+ ThreadUtils.shutdown(timer)
Review Comment:
Make the change in this thread. And make the executor a daemon thread.
I don't understand how it isn't applicable. The issue you identify is really
because the thread is not a daemon. That is what stops JVM exit, not the
execution of a task. The JVM exits even if a daemon thread is running.
Both parts of the change would be OK. But, maybe you want to revise the PR
description? because the problem and change here are not related to being
daemon.
--
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]
