hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jeffrey Zhong (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-9480) Regions are unexpectedly made offline in certain failure conditions
Date Wed, 11 Sep 2013 18:24:54 GMT

    [ https://issues.apache.org/jira/browse/HBASE-9480?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13764592#comment-13764592
] 

Jeffrey Zhong commented on HBASE-9480:
--------------------------------------

{quote}
If a region is indeed not served in a region server, I think we should delete the znode.
{quote}
The RIT zk node deletion here is on intermediate states(closing & closed) not the final
state nodes. Deleting them stops state transition abruptly and causes region lost as there
are multiple parties(master & region server) in state transition. We also see similar
RIT premature deletion(hbase-9387) cause region lost. IMHO, it's better to keep RIT znodes
and let region state machine proceed.
                
> Regions are unexpectedly made offline in certain failure conditions
> -------------------------------------------------------------------
>
>                 Key: HBASE-9480
>                 URL: https://issues.apache.org/jira/browse/HBASE-9480
>             Project: HBase
>          Issue Type: Bug
>            Reporter: Devaraj Das
>            Priority: Blocker
>             Fix For: 0.96.0
>
>         Attachments: 9480-1.txt, trunk-9480.patch
>
>
> Came across this issue (HBASE-9338 test):
> 1. Client issues a request to move a region from ServerA to ServerB
> 2. ServerA is compacting that region and doesn't close region immediately. In fact, it
takes a while to complete the request.
> 3. The master in the meantime, sends another close request.
> 4. ServerA sends it a NotServingRegionException
> 5. Master handles the exception, deletes the znode, and invokes regionOffline for the
said region.
> 6. ServerA fails to operate on ZK in the CloseRegionHandler since the node is deleted.
> The region is permanently offline.
> There are potentially other situations where when a RegionServer is offline and the client
asks for a region move off from that server, the master makes the region offline.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

Mime
View raw message