Github user mridulm commented on a diff in the pull request:
https://github.com/apache/spark/pull/17166#discussion_r107345386
--- Diff:
core/src/main/scala/org/apache/spark/scheduler/TaskSchedulerImpl.scala ---
@@ -467,7 +474,7 @@ private[spark] class TaskSchedulerImpl
private[scheduler](
taskState: TaskState,
reason: TaskFailedReason): Unit = synchronized {
taskSetManager.handleFailedTask(tid, taskState, reason)
- if (!taskSetManager.isZombie && taskState != TaskState.KILLED) {
+ if (!taskSetManager.isZombie) {
--- End diff --
First, the assumption that reviveOffers is inexpensive is incorrect.
Secondly, task kill happens for a few reasons automatically - and can cause
a bunch of tasks to be killed (for example, with speculative execution enabled,
when one task finishes, the speculative task will be killed).
This can cause a request storm of revive offers to be launched.
Given this, the change is incorrect and should be reverted.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
