You are viewing a plain text version of this content. The canonical link for it is here.

Posted to mapreduce-issues@hadoop.apache.org by "Eli Collins (JIRA)" <ji...@apache.org> on 2011/08/11 20:24:30 UTC

[jira] [Moved] (MAPREDUCE-2819) extend JobClient.runJob(JobConf) with the ability to take a timeout, so fail better during test runs

     [ https://issues.apache.org/jira/browse/MAPREDUCE-2819?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Eli Collins moved HADOOP-4639 to MAPREDUCE-2819:
------------------------------------------------

          Component/s:     (was: test)
                       test
    Affects Version/s:     (was: 0.20.0)
           Issue Type: Bug  (was: Improvement)
                  Key: MAPREDUCE-2819  (was: HADOOP-4639)
              Project: Hadoop Map/Reduce  (was: Hadoop Common)

> extend JobClient.runJob(JobConf) with the ability to take a timeout, so fail better during test runs
> ----------------------------------------------------------------------------------------------------
>
>                 Key: MAPREDUCE-2819
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-2819
>             Project: Hadoop Map/Reduce
>          Issue Type: Bug
>          Components: test
>            Reporter: Steve Loughran
>            Assignee: Steve Loughran
>            Priority: Minor
>
> Tests that submit jobs via JobClient hang until they are killed if something goes wrong in the back end -JobClient does not impose limits on how long runs should take, but JUnit does. If we had an overload of runJob() that took a timeout, JobClient could kill a job that was taking too long, extracting the stack trace and better diagnostics to the test reports. 

--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira