Github user jerryshao commented on a diff in the pull request:
https://github.com/apache/spark/pull/19580#discussion_r147325260
--- Diff:
core/src/main/scala/org/apache/spark/ExecutorAllocationManager.scala ---
@@ -267,6 +267,10 @@ private[spark] class ExecutorAllocationManager(
(numRunningOrPendingTasks + tasksPerExecutor - 1) / tasksPerExecutor
}
+ private def totalRunningTasks(): Int = synchronized {
--- End diff --
I'm not sure why do we need to add a method which only used for unit test.
If want to verify the behavior of `totalRunningTasks`, I think
`maxNumExecutorsNeeded` can also be used indirectly for verification.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
