[jira] [Commented] (TEZ-2404) Handle DataMovementEvent before its TaskAttemptCompletedEvent

["Jeff Zhang (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/TEZ-2404?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14526248#comment-14526248 ] 

Jeff Zhang commented on TEZ-2404:
---------------------------------

After rethinking, method 2 doesn't work. Because after TEZ-1897, VertexEvent/TaskEvent will be handled in different thread, it can't guarantee the order of TaskAttemptCompletedEvent & DataMovementEvent. So we have to use method 1 to route both DataMovementEvent and TaskAttemptCompletedEvent in Vertex, so that we can guarantee the ordering.  

> Handle DataMovementEvent before its TaskAttemptCompletedEvent
> -------------------------------------------------------------
>
>                 Key: TEZ-2404
>                 URL: https://issues.apache.org/jira/browse/TEZ-2404
>             Project: Apache Tez
>          Issue Type: Bug
>            Reporter: Jeff Zhang
>            Assignee: Jeff Zhang
>
> TEZ-2325 route TASK_ATTEMPT_COMPLETED_EVENT directly to the attempt, but it would cause recovery issue. Recovery need that DataMovement event is handled before TaskAttemptCompletedEvent, otherwise DataMovement event may be lost in recovering and cause the its dependent tasks hang.
> 2 Ways to fix this issue.
> 1. Still route TaskAtttemptCompletedEvent in Vertex
> 2. route DataMovementEvent before TaskAttemptCompeltedEvent in TezTaskAttemptListener



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)