hadoop-common-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Amareshwari Sriramadasu (JIRA)" <j...@apache.org>
Subject [jira] Commented: (HADOOP-4996) JobControl does not report killed jobs
Date Sat, 28 Feb 2009 04:09:13 GMT

    [ https://issues.apache.org/jira/browse/HADOOP-4996?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12677640#action_12677640

Amareshwari Sriramadasu commented on HADOOP-4996:

With the current code, if a job is FAILED/KILLED the JobControl shows the status as FAILED.
 If a dependent is FAILED/KILLED, the JobControl shows the status as DEPENDENT_FAILED. And
all these jobs are available in failedJobs list, which can be queried through getFailedJobs()

Is it that JobControl should also have new state KILLED and DEPENDENT_KILLED and a new api
getKilledJobs() for these jobs?
then, is it blocker?

> JobControl does not report killed jobs
> --------------------------------------
>                 Key: HADOOP-4996
>                 URL: https://issues.apache.org/jira/browse/HADOOP-4996
>             Project: Hadoop Core
>          Issue Type: Bug
>          Components: mapred
>    Affects Versions: 0.19.0
>            Reporter: Olga Natkovich
>            Assignee: Amareshwari Sriramadasu
>            Priority: Blocker
>             Fix For: 0.20.0
>         Attachments: patch-4996-testcase.txt
> After speaking with Arun and Owen, my understanding of the situation is that separate
killed job tracking was added in hadoop 18: http://issues.apache.org/jira/browse/HADOOP-3924.
> However, it does not look like this change was integrated into JobControl class. While
I have not verified this yet, it looks like, applications that use JobControl would no way
of knowing if one of the jobs was killed.
> This would be blocker for Pig to move to Hadoop 19.

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

View raw message