You are viewing a plain text version of this content. The canonical link for it is here.
Posted to reviews@spark.apache.org by GitBox <gi...@apache.org> on 2022/09/20 00:49:00 UTC

[GitHub] [spark] warrenzhu25 commented on a diff in pull request #37924: [SPARK-40481][CORE] Ignore stage fetch failure caused by decommissioned executor

warrenzhu25 commented on code in PR #37924:
URL: https://github.com/apache/spark/pull/37924#discussion_r974793372


##########
core/src/main/scala/org/apache/spark/scheduler/DAGScheduler.scala:
##########
@@ -1860,8 +1867,18 @@ private[spark] class DAGScheduler(
             s"(attempt ${failedStage.latestInfo.attemptNumber}) running")
         } else {
           failedStage.failedAttemptIds.add(task.stageAttemptId)
+          val ignoreStageFailure = ignoreDecommissionFetchFailure &&
+            isExecutorDecommissioned(taskScheduler, bmAddress)
+          if (ignoreStageFailure) {
+            logInfo("Ignoring fetch failure from $task of $failedStage attempt " +
+              s"${task.stageAttemptId} when count spark.stage.maxConsecutiveAttempts " +
+              "as executor ${bmAddress.executorId} is decommissioned and " +
+              s" ${config.STAGE_IGNORE_DECOMMISSION_FETCH_FAILURE.key}=true")
+          }
+
           val shouldAbortStage =
-            failedStage.failedAttemptIds.size >= maxConsecutiveStageAttempts ||
+            (!ignoreStageFailure &&
+              failedStage.failedAttemptIds.size >= maxConsecutiveStageAttempts) ||
             disallowStageRetryForTest

Review Comment:
   Good catch. Updated.



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org