Github user JoshRosen commented on a diff in the pull request: https://github.com/apache/spark/pull/14931#discussion_r77745305 --- Diff: core/src/main/scala/org/apache/spark/scheduler/cluster/StandaloneSchedulerBackend.scala --- @@ -153,7 +153,7 @@ private[spark] class StandaloneSchedulerBackend( override def executorRemoved(fullId: String, message: String, exitStatus: Option[Int]) { val reason: ExecutorLossReason = exitStatus match { case Some(code) => ExecutorExited(code, exitCausedByApp = true, message) - case None => SlaveLost(message) + case None => SlaveLost(message, workerLost = true /* worker loss event from master */) --- End diff -- This assumes that `exitStatus == None` implies that a worker was lost, but there are some corner-cases where this isn't necessarily true (e.g. if an executor kill fails). Looking through both the 1.6.x and 2.0.x code, it appears that `ExecutorStatus.LOST` is used exclusively for denoting whole-worker-loss, so I think that we should check that status here instead of assuming `true`. Other than that minor corner-case, this patch looks good to me, so I'll merge once we fix this.
--- 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