hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "zhou wenjian (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-6391) Master restart when enabling table will lead to region assignned twice
Date Mon, 16 Jul 2012 09:15:35 GMT

    [ https://issues.apache.org/jira/browse/HBASE-6391?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13414946#comment-13414946

zhou wenjian commented on HBASE-6391:

thanks for the reply.
I mistake for the use of checkIfRegionsBelongsToEnabling.

in your patch for HBASE-6317
If there exists enabling state in zk and restart the whole cluster, it will be considered
as a failover, will this slow down the startup of master?  Can't we got the explicit range
of what cause double assign or inconsistent

Correct me if i'm wrong
> Master restart when enabling table will lead to region assignned twice
> ----------------------------------------------------------------------
>                 Key: HBASE-6391
>                 URL: https://issues.apache.org/jira/browse/HBASE-6391
>             Project: HBase
>          Issue Type: Bug
>          Components: master
>    Affects Versions: 0.94.0
>            Reporter: zhou wenjian
>             Fix For: 0.94.2
> The Scenario can be reproduce below.
> Enabling an table, some region is online on regionserver,some are still being processed.
> And restart the master.
> when master failover:
>         // Region is being served and on an active server
>         // add only if region not in disabled and enabling table
>         if (false == checkIfRegionBelongsToDisabled(regionInfo)
>             && false == checkIfRegionsBelongsToEnabling(regionInfo)) {
>           regions.put(regionInfo, regionLocation);
>           addToServers(regionLocation, regionInfo);
>         }
> the opened region will not add to the Regions in master.
> and in the following recoverTableInEnablingState,the region will be assigned again.
> that will lead to the cluster inconsistent

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


View raw message