hadoop-common-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Steve Loughran (JIRA)" <j...@apache.org>
Subject [jira] Commented: (HADOOP-4374) JVM should not be killed but given an opportunity to exit gracefully
Date Mon, 30 Mar 2009 13:14:50 GMT

    [ https://issues.apache.org/jira/browse/HADOOP-4374?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12693757#action_12693757
] 

Steve Loughran commented on HADOOP-4374:
----------------------------------------

Looks good to me. 


One question for Amareshwari: why log the exception with stringifyException instead of log.error("Exception
" + e, e) ? That latter one retains the stack trace and the like for longer, and lets you
decide what to do with the details in the logging configuration. I know a lot of the code
uses stringifyException(), but thought the codebase was moving away from it?

> JVM should not be killed but given an opportunity to exit gracefully
> --------------------------------------------------------------------
>
>                 Key: HADOOP-4374
>                 URL: https://issues.apache.org/jira/browse/HADOOP-4374
>             Project: Hadoop Core
>          Issue Type: Bug
>          Components: mapred
>    Affects Versions: 0.19.0
>            Reporter: Devaraj Das
>            Assignee: Ravi Gummadi
>             Fix For: 0.20.0
>
>         Attachments: HADOOP-4374.patch, HADOOP-4374.v1.1.patch, HADOOP-4374.v1.patch
>
>
> When the tasktracker picks an idle JVM for purging, it should signal the JVM to exit
gracefully, rather than forcefully killing it. This might have the unfortunate side effect
of logs not fully flushed yet in some cases.

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


Mime
View raw message