hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Uma Maheswara Rao G (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-7878) recoverFileLease does not check return value of recoverLease
Date Wed, 27 Feb 2013 05:24:15 GMT

    [ https://issues.apache.org/jira/browse/HBASE-7878?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13588013#comment-13588013
] 

Uma Maheswara Rao G commented on HBASE-7878:
--------------------------------------------

+1 for committing this. 
This test can ensure to do recoverlease with append call. But practically I don't see any
reason for failing recoverLease call and passing append call.
But we don't want to remove append call, I think this is the way to test it.
Another negative test could be, call the recoverLease on deleted non existent WAL file. There
also recoverLease call will fail. Then call will go to append, here append also will be failed
because fileDoesNotexist. and it should not hand for long time in such case.

I found one case, that if there are no blocks allocated but file is just created. then recoverLease
call actually will finalize the inode and return true internally. But that value will not
be really propagated to out. So, recoverLease will return false to out side. That should not
be a problem because, next recoverLease call will get true because inode already finalized.
If possible may be we can have one test, but not forcing to add here. 

Thanks Ted, you can go ahead and commit if Lars also ok for it.
                
> recoverFileLease does not check return value of recoverLease
> ------------------------------------------------------------
>
>                 Key: HBASE-7878
>                 URL: https://issues.apache.org/jira/browse/HBASE-7878
>             Project: HBase
>          Issue Type: Bug
>          Components: util
>    Affects Versions: 0.95.0, 0.94.6
>            Reporter: Eric Newton
>            Assignee: Ted Yu
>            Priority: Critical
>             Fix For: 0.95.0, 0.98.0, 0.94.6
>
>         Attachments: 7878-trunk-v2.txt, 7878-trunk-v3.txt, 7878-trunk-v4.txt, 7878-trunk-v5.txt
>
>
> I think this is a problem, so I'm opening a ticket so an HBase person takes a look.
> Apache Accumulo has moved its write-ahead log to HDFS. I modeled the lease recovery for
Accumulo after HBase's lease recovery.  During testing, we experienced data loss.  I found
it is necessary to wait until recoverLease returns true to know that the file has been truly
closed.  In FSHDFSUtils, the return result of recoverLease is not checked. In the unit tests
created to check lease recovery in HBASE-2645, the return result of recoverLease is always
checked.
> I think FSHDFSUtils should be modified to check the return result, and wait until it
returns true.

--
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