[
https://issues.apache.org/jira/browse/SPARK-19698?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15881644#comment-15881644
]
Jisoo Kim commented on SPARK-19698:
-----------------------------------
[~kayousterhout] If the failed task gets re-tried, as long as Driver doesn't
shut down before the next attempt finishes, it should be ok because the next
attempt will upload a file as intended. That's actually similar to what
happened in my workload, executor was lost due to OOME and stage was
resubmitted eventually. If the driver didn't think that the job was done,
things would've been fine. The driver didn't mark the partition that the failed
task was responsible for as "finished", so in the next attempt, the task
finished successfully (and there were no race condition for this specific task
because the executor that was running this task was lost) but one of the other
tasks had a such problem. One thing I am not sure about my solution is a
possible performance regression, but I think it might be better than having
some kind of an "incorrect" external state unless it is not recommended and not
a good practice to have a task to modify some external state.
> Race condition in stale attempt task completion vs current attempt task
> completion when task is doing persistent state changes
> ------------------------------------------------------------------------------------------------------------------------------
>
> Key: SPARK-19698
> URL: https://issues.apache.org/jira/browse/SPARK-19698
> Project: Spark
> Issue Type: Bug
> Components: Mesos, Spark Core
> Affects Versions: 2.0.0
> Reporter: Charles Allen
>
> We have encountered a strange scenario in our production environment. Below
> is the best guess we have right now as to what's going on.
> Potentially, the final stage of a job has a failure in one of the tasks (such
> as OOME on the executor) which can cause tasks for that stage to be
> relaunched in a second attempt.
> https://github.com/apache/spark/blob/v2.1.0/core/src/main/scala/org/apache/spark/scheduler/DAGScheduler.scala#L1155
> keeps track of which tasks have been completed, but does NOT keep track of
> which attempt those tasks were completed in. As such, we have encountered a
> scenario where a particular task gets executed twice in different stage
> attempts, and the DAGScheduler does not consider if the second attempt is
> still running. This means if the first task attempt succeeded, the second
> attempt can be cancelled part-way through its run cycle if all other tasks
> (including the prior failed) are completed successfully.
> What this means is that if a task is manipulating some state somewhere (for
> example: a upload-to-temporary-file-location, then delete-then-move on an
> underlying s3n storage implementation) the driver can improperly shutdown the
> running (2nd attempt) task between state manipulations, leaving the
> persistent state in a bad state since the 2nd attempt never got to complete
> its manipulations, and was terminated prematurely at some arbitrary point in
> its state change logic (ex: finished the delete but not the move).
> This is using the mesos coarse grained executor. It is unclear if this
> behavior is limited to the mesos coarse grained executor or not.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
