flink-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "ASF GitHub Bot (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (FLINK-3800) ExecutionGraphs can become orphans
Date Thu, 23 Jun 2016 13:10:16 GMT

    [ https://issues.apache.org/jira/browse/FLINK-3800?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15346399#comment-15346399
] 

ASF GitHub Bot commented on FLINK-3800:
---------------------------------------

Github user tillrohrmann commented on the issue:

    https://github.com/apache/flink/pull/2096
  
    Thanks for your thorough review @uce.
    
    - We have talked offline concerning the status in the web interface and it turned out
to be not a problem since the jobs are directly removed from `currentJobs` in the `JobManager`
when `cancelAndClearEverything` is called. This implies that the jobs will no longer be shown
in the web interface.
    
    - I agree that it is a good idea to replace the `SuppressRestartException` by a mechanism
to disable the `RestartStrategies`. I've opened a [JIRA issue](https://issues.apache.org/jira/browse/FLINK-4112)
to keep track of the effort. I think, however, that the mechanism should be implemented as
part of the work on this issue. That way we won't mix pull requests with each other.
    
    - You're right. I've addressed the comment and added `SUSPENDED` to the set of states
which don't throw an exception when encountered in `ExecutionGraph.restart`.
    
    - You're right concerning the `JobManagerProcess` tests. Maybe we can refactor some of
them in the future by applying a similar pattern as it was used in `JobManagerHARecoveryTest`.


> ExecutionGraphs can become orphans
> ----------------------------------
>
>                 Key: FLINK-3800
>                 URL: https://issues.apache.org/jira/browse/FLINK-3800
>             Project: Flink
>          Issue Type: Bug
>          Components: JobManager
>    Affects Versions: 1.0.0, 1.1.0
>            Reporter: Till Rohrmann
>            Assignee: Till Rohrmann
>
> The {{JobManager.cancelAndClearEverything}} method fails all currently executed jobs
on the {{JobManager}} and then clears the list of {{currentJobs}} kept in the JobManager.
This can become problematic if the user has set a restart strategy for a job, because the
{{RestartStrategy}} will try to restart the job. This can lead to unwanted re-deployments
of the job which consumes resources and thus will trouble the execution of other jobs. If
the restart strategy never stops, then this prevents that the {{ExecutionGraph}} from ever
being properly terminated.



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

Mime
View raw message