[ 
https://issues.apache.org/jira/browse/BEAM-1594?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15893219#comment-15893219
 ] 

ASF GitHub Bot commented on BEAM-1594:
--------------------------------------

GitHub user jkff opened a pull request:

    https://github.com/apache/beam/pull/2146

    [BEAM-1594] Supports DRAINING and DRAINED states

    Dataflow's DRAINING is treated as RUNNING, and DRAINED as DONE.
    
    R: @tgroh 
    CC: @dhalperi 

You can merge this pull request into a Git repository by running:

    $ git pull https://github.com/jkff/incubator-beam df-drained

Alternatively you can review and apply these changes as the patch at:

    https://github.com/apache/beam/pull/2146.patch

To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:

    This closes #2146
    
----
commit be64925e3c0095d49760a84583677fdf54e47ef6
Author: Eugene Kirpichov <[email protected]>
Date:   2017-03-02T23:12:14Z

    [BEAM-1594] Supports DRAINING and DRAINED states
    
    Dataflow's DRAINING is treated as RUNNING, and DRAINED as DONE.

----


> JOB_STATE_DRAINED should be terminal for waitUntilFinish()
> ----------------------------------------------------------
>
>                 Key: BEAM-1594
>                 URL: https://issues.apache.org/jira/browse/BEAM-1594
>             Project: Beam
>          Issue Type: Bug
>          Components: runner-dataflow
>            Reporter: Eugene Kirpichov
>            Assignee: Eugene Kirpichov
>
> Dataflow runner supports pipeline draining for streaming pipelines, which 
> stops consuming input and waits for all watermarks to progress to infinity 
> and then declares the pipeline complete. When watermarks progress to infinity 
> after a drain request, the Dataflow runner will mark the job DRAINED (In the 
> future, when watermarks progress to infinity without a drain request, it will 
> mark the job DONE instead). JOB_STATE_DRAINED is logically a terminal state - 
> no more data will be processed and the pipeline's resources are torn down - 
> and waitUntilFinish() should terminate if the job enters this state.



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)

Reply via email to