Github user mateiz commented on a diff in the pull request:
https://github.com/apache/spark/pull/186#discussion_r11930437
--- Diff: core/src/main/scala/org/apache/spark/scheduler/DAGScheduler.scala
---
@@ -730,6 +588,100 @@ class DAGScheduler(
}
}
+ private[scheduler] def handleJobGroupCancelled(groupId: String) {
+ // Cancel all jobs belonging to this job group.
+ // First finds all active jobs with this group id, and then kill
stages for them.
+ val activeInGroup = activeJobs.filter(activeJob =>
+ groupId == activeJob.properties.get(SparkContext.SPARK_JOB_GROUP_ID))
+ val jobIds = activeInGroup.map(_.jobId)
+ jobIds.foreach(handleJobCancellation(_, "part of cancel job group"))
+ submitWaitingStages()
+ }
+
+ private[scheduler] def handleBeginEvent(task: Task[_], taskInfo:
TaskInfo) {
+ for (stage <- stageIdToStage.get(task.stageId); stageInfo <-
stageToInfos.get(stage)) {
+ if (taskInfo.serializedSize > DAGScheduler.TASK_SIZE_TO_WARN * 1024
&&
+ !stageInfo.emittedTaskSizeWarning) {
+ stageInfo.emittedTaskSizeWarning = true
+ logWarning(("Stage %d (%s) contains a task of very large " +
+ "size (%d KB). The maximum recommended task size is %d
KB.").format(
+ task.stageId, stageInfo.name, taskInfo.serializedSize / 1024,
+ DAGScheduler.TASK_SIZE_TO_WARN))
+ }
+ }
+ listenerBus.post(SparkListenerTaskStart(task.stageId, taskInfo))
+ submitWaitingStages()
--- End diff --
Ah I see, before we were submitting after all events in the match. I find
this somewhat confusing but I guess we have to do it this way.
---
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.
---
