hbase-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jean-Daniel Cryans (JIRA)" <j...@apache.org>
Subject [jira] Commented: (HBASE-845) disable then drop table is messy when table is > 4 or 5 M rows
Date Wed, 28 Jan 2009 21:08:59 GMT

    [ https://issues.apache.org/jira/browse/HBASE-845?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12668162#action_12668162
] 

Jean-Daniel Cryans commented on HBASE-845:
------------------------------------------

I tested my patch on a 79 regions table with well over 10M rows and isTableEnabled reliably
tells me if all regions are disabled or not. But this is a new problem, if half of the regions
are disabled, it tells us that it is enabled....

> disable then drop table is messy when table is > 4 or 5 M rows
> --------------------------------------------------------------
>
>                 Key: HBASE-845
>                 URL: https://issues.apache.org/jira/browse/HBASE-845
>             Project: Hadoop HBase
>          Issue Type: Bug
>            Reporter: stack
>         Attachments: hbase-845.patch
>
>
> In current trunk, if i load a table of 8M rows and then try and delete it, the disable
returns saying the table was successfully deleted but when I then try to drop the table, it
says table not disabled.  I run the disable/drop cycle a few more times and still fails. 
Eventually, if I wait long enough, it succeeds.   Maybe the table drop should just block if
table is seen to have disabled regions in it.  As is, its a little disorientating the way
it works.  Could lead admins to distrust status messages emitted.

-- 
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