You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@hbase.apache.org by "Nate Putnam (JIRA)" <ji...@apache.org> on 2011/09/08 23:43:09 UTC

[jira] [Created] (HBASE-4356) Wide rows can cause OOME when compacting

Wide rows can cause OOME when compacting
----------------------------------------

                 Key: HBASE-4356
                 URL: https://issues.apache.org/jira/browse/HBASE-4356
             Project: HBase
          Issue Type: Bug
    Affects Versions: 0.90.4
            Reporter: Nate Putnam


The scanner used for compaction doesn't limit the number of columns retrieved when doing a compaction. If a row exists with tens of millions of rows it can fill all available memory and crash the region server. 

It would be better if the scanner could page through the columns of wide rows when performing a compaction. 

--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

[jira] [Commented] (HBASE-4356) Wide rows can cause OOME when compacting

Posted by "Nate Putnam (JIRA)" <ji...@apache.org>.
    [ https://issues.apache.org/jira/browse/HBASE-4356?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13100772#comment-13100772 ] 

Nate Putnam commented on HBASE-4356:
------------------------------------

Yep. Looks like the same issue. I'll close this one. 

> Wide rows can cause OOME when compacting
> ----------------------------------------
>
>                 Key: HBASE-4356
>                 URL: https://issues.apache.org/jira/browse/HBASE-4356
>             Project: HBase
>          Issue Type: Bug
>    Affects Versions: 0.90.4
>            Reporter: Nate Putnam
>
> The scanner used for compaction doesn't limit the number of columns retrieved when doing a compaction. If a row exists with tens of millions of rows it can fill all available memory and crash the region server. 
> It would be better if the scanner could page through the columns of wide rows when performing a compaction. 

--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

[jira] [Commented] (HBASE-4356) Wide rows can cause OOME when compacting

Posted by "Todd Lipcon (JIRA)" <ji...@apache.org>.
    [ https://issues.apache.org/jira/browse/HBASE-4356?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13100723#comment-13100723 ] 

Todd Lipcon commented on HBASE-4356:
------------------------------------

Dup of HBASE-3421?

> Wide rows can cause OOME when compacting
> ----------------------------------------
>
>                 Key: HBASE-4356
>                 URL: https://issues.apache.org/jira/browse/HBASE-4356
>             Project: HBase
>          Issue Type: Bug
>    Affects Versions: 0.90.4
>            Reporter: Nate Putnam
>
> The scanner used for compaction doesn't limit the number of columns retrieved when doing a compaction. If a row exists with tens of millions of rows it can fill all available memory and crash the region server. 
> It would be better if the scanner could page through the columns of wide rows when performing a compaction. 

--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

[jira] [Resolved] (HBASE-4356) Wide rows can cause OOME when compacting

Posted by "Nate Putnam (JIRA)" <ji...@apache.org>.
     [ https://issues.apache.org/jira/browse/HBASE-4356?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Nate Putnam resolved HBASE-4356.
--------------------------------

    Resolution: Duplicate

Dup of HBAE-3421

> Wide rows can cause OOME when compacting
> ----------------------------------------
>
>                 Key: HBASE-4356
>                 URL: https://issues.apache.org/jira/browse/HBASE-4356
>             Project: HBase
>          Issue Type: Bug
>    Affects Versions: 0.90.4
>            Reporter: Nate Putnam
>
> The scanner used for compaction doesn't limit the number of columns retrieved when doing a compaction. If a row exists with tens of millions of rows it can fill all available memory and crash the region server. 
> It would be better if the scanner could page through the columns of wide rows when performing a compaction. 

--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira