Github user vanzin commented on a diff in the pull request: https://github.com/apache/spark/pull/17480#discussion_r111204631 --- Diff: core/src/main/scala/org/apache/spark/ExecutorAllocationManager.scala --- @@ -249,7 +249,14 @@ private[spark] class ExecutorAllocationManager( * yarn-client mode when AM re-registers after a failure. */ def reset(): Unit = synchronized { - initializing = true + /** + * When some tasks need to be scheduled and initial executor = 0, resetting the initializing + * field may cause it to not be set to false in yarn. + * SPARK-20079: https://issues.apache.org/jira/browse/SPARK-20079 + */ + if (maxNumExecutorsNeeded() == 0) { + initializing = true --- End diff -- This is the comment attached to `initializing`: ``` // Whether we are still waiting for the initial set of executors to be allocated. // While this is true, we will not cancel outstanding executor requests. This is // set to false when: // (1) a stage is submitted, or // (2) an executor idle timeout has elapsed. ``` None of those situations are true in reset; we're not waiting for the initial set of executors anymore. A stage might have been submitted already and flipped the value of `initializing`. Many other things have happened. So I still don't understand why is it /ever/ correct to set initializing to false in `reset`.
--- 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 infrastruct...@apache.org or file a JIRA ticket with INFRA. --- --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org