You are viewing a plain text version of this content. The canonical link for it is here.
Posted to common-dev@hadoop.apache.org by "Hairong Kuang (JIRA)" <ji...@apache.org> on 2006/05/27 01:07:29 UTC

[jira] Created: (HADOOP-257) starting one data node thread to manage multiple data directories

starting one data node thread to manage multiple data directories
-----------------------------------------------------------------

         Key: HADOOP-257
         URL: http://issues.apache.org/jira/browse/HADOOP-257
     Project: Hadoop
        Type: Improvement

  Components: dfs  
    Reporter: Hairong Kuang


If a data node is configured with multiple data directories, current implementation of dfs will start multiple data node threads, each of which manages one data directory and talks to its name node independently. From the name node's point of view, it sees multiple data nodes instead of one.

I feel that a more scalable solution should be to start one data node thread that manages multiple data diretories. But the one data node thread needs to take care of the block allocation problem.

-- 
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators:
   http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see:
   http://www.atlassian.com/software/jira


[jira] Updated: (HADOOP-257) starting one data node thread to manage multiple data directories

Posted by "Sameer Paranjpye (JIRA)" <ji...@apache.org>.
     [ http://issues.apache.org/jira/browse/HADOOP-257?page=all ]

Sameer Paranjpye updated HADOOP-257:
------------------------------------

    Fix Version: 0.4
        Version: 0.2
      Assign To: Konstantin Shvachko

> starting one data node thread to manage multiple data directories
> -----------------------------------------------------------------
>
>          Key: HADOOP-257
>          URL: http://issues.apache.org/jira/browse/HADOOP-257
>      Project: Hadoop
>         Type: Improvement

>   Components: dfs
>     Versions: 0.2
>     Reporter: Hairong Kuang
>     Assignee: Konstantin Shvachko
>      Fix For: 0.4

>
> If a data node is configured with multiple data directories, current implementation of dfs will start multiple data node threads, each of which manages one data directory and talks to its name node independently. From the name node's point of view, it sees multiple data nodes instead of one.
> I feel that a more scalable solution should be to start one data node thread that manages multiple data diretories. But the one data node thread needs to take care of the block allocation problem.

-- 
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators:
   http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see:
   http://www.atlassian.com/software/jira


[jira] Resolved: (HADOOP-257) starting one data node thread to manage multiple data directories

Posted by "Sameer Paranjpye (JIRA)" <ji...@apache.org>.
     [ http://issues.apache.org/jira/browse/HADOOP-257?page=all ]
     
Sameer Paranjpye resolved HADOOP-257:
-------------------------------------

    Resolution: Duplicate

Duplicate of HADOOP-64

> starting one data node thread to manage multiple data directories
> -----------------------------------------------------------------
>
>          Key: HADOOP-257
>          URL: http://issues.apache.org/jira/browse/HADOOP-257
>      Project: Hadoop
>         Type: Improvement

>   Components: dfs
>     Versions: 0.2
>     Reporter: Hairong Kuang
>     Assignee: Konstantin Shvachko
>      Fix For: 0.4

>
> If a data node is configured with multiple data directories, current implementation of dfs will start multiple data node threads, each of which manages one data directory and talks to its name node independently. From the name node's point of view, it sees multiple data nodes instead of one.
> I feel that a more scalable solution should be to start one data node thread that manages multiple data diretories. But the one data node thread needs to take care of the block allocation problem.

-- 
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators:
   http://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see:
   http://www.atlassian.com/software/jira