Mailing-List: contact issues-help@hbase.apache.org; run by ezmlm
Precedence: bulk
Date: Fri, 6 Jul 2012 17:45:34 +0000 (UTC)
From: "ramkrishna.s.vasudevan (JIRA)" <jira@apache.org>
To: issues@hbase.apache.org
Message-ID: <778178522.15082.1341596734800.JavaMail.jiratomcat@issues-vm>
In-Reply-To: <689805798.12804.1341554374837.JavaMail.jiratomcat@issues-vm>
Subject: [jira] [Commented] (HBASE-6336) Split point should not be equal
 with start row or end row
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


    [ https://issues.apache.org/jira/browse/HBASE-6336?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13408176#comment-13408176 ] 

ramkrishna.s.vasudevan commented on HBASE-6336:
-----------------------------------------------

I think this is fine.
One more thing we noticed here is Suppose
I have a table with no split keys.  And i have very few data in that.  Now if i try to split we split from 
[] to [firstrow]
[firstrow] to []
Why we get first row as 1st region's end key is i have only one root index and it will select the first row as the midkey.  
But here in the first region there are no kvs at all and hence we flush an empty file.  Now when we again call split passing the table name for the first region we get an ioexception from
{code}
if (rootCount == 0)
        throw new IOException("HFile empty");
{code}
It may not be a bug but just found this.
                
> Split point should not be equal with start row or end row
> ---------------------------------------------------------
>
>                 Key: HBASE-6336
>                 URL: https://issues.apache.org/jira/browse/HBASE-6336
>             Project: HBase
>          Issue Type: Bug
>          Components: regionserver
>            Reporter: chunhui shen
>            Assignee: chunhui shen
>             Fix For: 0.96.0
>
>         Attachments: HBASE-6336.patch
>
>
> Should we allow split point equal with region's start row or end row?
> {code}
> // if the midkey is the same as the first and last keys, then we cannot
>         // (ever) split this region.
>         if (this.comparator.compareRows(mk, firstKey) == 0 &&
>             this.comparator.compareRows(mk, lastKey) == 0) {
>           if (LOG.isDebugEnabled()) {
>             LOG.debug("cannot split because midkey is the same as first or " +
>               "last row");
>           }
> {code}
> Here, I think it is a mistake.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira