You are viewing a plain text version of this content. The canonical link for it is here.

Posted to dev@hbase.apache.org by "Doğacan Güney (JIRA)" <ji...@apache.org> on 2009/07/10 23:27:15 UTC

[jira] Created: (HBASE-1644) Result.row is cached in getRow; this breaks MapReduce

Result.row is cached in getRow; this breaks MapReduce
-----------------------------------------------------

                 Key: HBASE-1644
                 URL: https://issues.apache.org/jira/browse/HBASE-1644
             Project: Hadoop HBase
          Issue Type: Bug
    Affects Versions: 0.20.0
            Reporter: Doğacan Güney


In Result#getRow row field is computed (if row is null) and then is cached for further uses. But since MapReduce uses the same Result instance through different map()/reduce() calls, row field is not updated when Result instance changes.

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

[jira] Resolved: (HBASE-1644) Result.row is cached in getRow; this breaks MapReduce

Posted by "stack (JIRA)" <ji...@apache.org>.

     [ https://issues.apache.org/jira/browse/HBASE-1644?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

stack resolved HBASE-1644.
--------------------------

       Resolution: Fixed
    Fix Version/s: 0.20.0
     Hadoop Flags: [Reviewed]

Committed.  Thanks for the patch Doğacan.

> Result.row is cached in getRow; this breaks MapReduce
> -----------------------------------------------------
>
>                 Key: HBASE-1644
>                 URL: https://issues.apache.org/jira/browse/HBASE-1644
>             Project: Hadoop HBase
>          Issue Type: Bug
>    Affects Versions: 0.20.0
>            Reporter: Doğacan Güney
>             Fix For: 0.20.0
>
>         Attachments: rownull.patch
>
>
> In Result#getRow row field is computed (if row is null) and then is cached for further uses. But since MapReduce uses the same Result instance through different map()/reduce() calls, row field is not updated when Result instance changes.

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

[jira] Updated: (HBASE-1644) Result.row is cached in getRow; this breaks MapReduce

Posted by "Doğacan Güney (JIRA)" <ji...@apache.org>.

     [ https://issues.apache.org/jira/browse/HBASE-1644?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Doğacan Güney updated HBASE-1644:
---------------------------------

    Attachment: rownull.patch

Patch for the problem.

Also TableInputFormatBase#nextKeyValue reads Result into a temporary result variable then copies it into value field. AFAICS, there is no point in doing that. So this patch also removes the local result variable.

> Result.row is cached in getRow; this breaks MapReduce
> -----------------------------------------------------
>
>                 Key: HBASE-1644
>                 URL: https://issues.apache.org/jira/browse/HBASE-1644
>             Project: Hadoop HBase
>          Issue Type: Bug
>    Affects Versions: 0.20.0
>            Reporter: Doğacan Güney
>         Attachments: rownull.patch
>
>
> In Result#getRow row field is computed (if row is null) and then is cached for further uses. But since MapReduce uses the same Result instance through different map()/reduce() calls, row field is not updated when Result instance changes.

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