Github user GraceH commented on a diff in the pull request:
https://github.com/apache/spark/pull/9796#discussion_r45283688
--- Diff:
core/src/main/scala/org/apache/spark/ExecutorAllocationManager.scala ---
@@ -408,7 +408,8 @@ private[spark] class ExecutorAllocationManager(
executorsPendingToRemove.add(executorId)
true
} else {
- logWarning(s"Unable to reach the cluster manager to kill executor
$executorId!")
+ logWarning(s"Unable to reach the cluster manager to kill executor
$executorId!" +
+ s"Or no executor to kill!")
--- End diff --
1. or maybe we need to change that to "or executor is not eligible to
kill". since there are two situations to get empty list: a) busy b) kill that
repeatedly
2. There is not action to take in other parts of the code. The problem here
is if we not change the API (return type). it is really hard to tell it is not
acknowledged or no executor to kill. If it is not acknowledged, we should print
warning, but if no executor to kill, it is the normal case with info level.
---
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]
