hadoop-mapreduce-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Konstantin Boudnik (JIRA)" <j...@apache.org>
Subject [jira] Commented: (MAPREDUCE-1827) Task Killing/Failing tests for a streaming job.
Date Wed, 02 Jun 2010 18:47:40 GMT

    [ https://issues.apache.org/jira/browse/MAPREDUCE-1827?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12874735#action_12874735

Konstantin Boudnik commented on MAPREDUCE-1827:

- {{@AfterClass}} method calls {{cluster.tearDown()}} twice. It is really necessary?
- there is a typo
+        "seelp time of 3 secs.", counter > 3 && TaskStatus.State.KILLED ==
- in the {{JobID getJobIdOfRunningStreamJob(String [] runtimeArgs) }} you use {{inputDir}}
and {{outputDir}} only once. There's no point to have a special variables to store their String
representation for a single use.
- it is a good idea to declare objects you aren't going to modify as {{final}}. I.e.
+    final RunStreamingJob streamJobThread = new RunStreamingJob(conf,
+        streamJob,streamingArgs);
- there's no need to assign {{null]] to a variable and then immediately assign a real object
 to it. It's clearer (from the code standpoint) and less clogged (from the logic flow point
of view)
- does {{System.getProperty("user.dir")}} points to {{src/contrib/streaming}} during the test
execution because otherwise {[/src/test/system/scripts/StreamMapper.sh}} will be looking for
the script in a wrong place. According to {{build-contrib.xml}} 
      <sysproperty key="user.dir" value="${build.test}/data"/>
so please make sure your assumptions are correct.

> Task Killing/Failing tests for a streaming job.
> -----------------------------------------------
>                 Key: MAPREDUCE-1827
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-1827
>             Project: Hadoop Map/Reduce
>          Issue Type: Task
>          Components: test
>            Reporter: Vinay Kumar Thota
>            Assignee: Vinay Kumar Thota
>         Attachments: 1827-ydist-security.patch
> 1. Set the sleep time for the tasks is 3 seconds and kill the task of streaming job using
SIGKILL. After that  verify whether task is killed after 3 seconds or not and also verify
whether job is succeeded or not.
> 2. Set the maximum attempts for the maps and reducers are one. make the task to fail
and verify whether task  is failed or not.Also verify whether the job is failed or not.

This message is automatically generated by JIRA.
You can reply to this email to add a comment to the issue online.

View raw message