hadoop-common-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Tsz Wo (Nicholas), SZE (JIRA)" <j...@apache.org>
Subject [jira] Updated: (HADOOP-5564) hadoop command uses large JVM heap size
Date Tue, 24 Mar 2009 18:50:50 GMT

     [ https://issues.apache.org/jira/browse/HADOOP-5564?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Tsz Wo (Nicholas), SZE updated HADOOP-5564:
-------------------------------------------

    Hadoop Flags: [Reviewed]

+1 patch looks good

Talked to Koji and Suresh offline.  We have considered using -client.  However, the [tool
doc|http://java.sun.com/javase/6/docs/technotes/tools/solaris/java.html] says that the -client
option may be ignored by the jdk.

> hadoop command uses large JVM heap size
> ---------------------------------------
>
>                 Key: HADOOP-5564
>                 URL: https://issues.apache.org/jira/browse/HADOOP-5564
>             Project: Hadoop Core
>          Issue Type: Bug
>          Components: build
>            Reporter: Suresh Srinivas
>            Assignee: Suresh Srinivas
>            Priority: Blocker
>             Fix For: 0.20.0
>
>         Attachments: 5564.patch
>
>
> Command used to determine JAVA_PLATFORM in bin/hadoop command does not set the heap size.
The command uses default 1GB heap size. The tasks invoking hadoop command end up  using large
heap size in streaming jobs. If the maximum memory that can be used by a task is restricted,
this could result in map/reduce job failures.

-- 
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