hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Gary Helmling (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-8978) Restore TestLogRollAbort upon review
Date Tue, 23 Jul 2013 07:16:52 GMT

    [ https://issues.apache.org/jira/browse/HBASE-8978?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13716186#comment-13716186

Gary Helmling commented on HBASE-8978:

After a lot of trying, I still haven't been able to reproduce this hang.  All told, I've run
TestLogRollAbort 300+ times using Hadoop 1 and 2.0 profiles, but no hang.  Regarding the comments
on HBASE-8939, from what I can tell, we should not be hanging on doing the join with the LogSyncer
thread either.  On triggering LogSyncer.close(), it looks like the LogSyncer thread will attempt
one more sync() (which should fail) and then exit.

So I'm not seeing where the hang is coming in.  Could it be an earlier test causing the hang
> Restore TestLogRollAbort upon review
> ------------------------------------
>                 Key: HBASE-8978
>                 URL: https://issues.apache.org/jira/browse/HBASE-8978
>             Project: HBase
>          Issue Type: Task
>          Components: test
>            Reporter: stack
>            Assignee: Gary Helmling
>             Fix For: 0.95.2
> Gary, would you mind taking a look?  No rush.  
> The test was added by HBASE-4282.  It is a good test against data loss but it is around
deferred flush which has changed a bunch since test was originally written.
> I removed it over in hbase-8977 because I caught it hanging and breaking our build (hbase-8939).
 If you run it, you'll see it can't shutdown.  We are bound up waiting on a regionserver to

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

View raw message