hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Todd Lipcon (JIRA)" <j...@apache.org>
Subject [jira] Assigned: (HBASE-2757) Regions not assigned after HBASE-2694 went in
Date Mon, 21 Jun 2010 00:46:24 GMT

     [ https://issues.apache.org/jira/browse/HBASE-2757?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Todd Lipcon reassigned HBASE-2757:
----------------------------------

    Assignee: Todd Lipcon

> Regions not assigned after HBASE-2694 went in
> ---------------------------------------------
>
>                 Key: HBASE-2757
>                 URL: https://issues.apache.org/jira/browse/HBASE-2757
>             Project: HBase
>          Issue Type: Bug
>            Reporter: stack
>            Assignee: Todd Lipcon
>            Priority: Blocker
>             Fix For: 0.21.0
>
>         Attachments: hbase-2757.txt
>
>
> Looking into the Hudson failure, http://hudson.zones.apache.org/hudson/view/HBase/job/HBase-TRUNK/1339/
at the failing org.apache.hadoop.hbase.client.TestFromClientSide.testRegionCachePreWarm test,
I seem to be seeing a case of regions being added to UNASSIGNED up in zk but then subsequently
nothing.   Adjacent regions are being similarily added but these are being assigned out. 
Somethings up.  It seems to be causing the above test failure.
> Its hard to see in the logs.... 
> Grep for this line:
>     LOG.info("Starting testRegionCachePreWarm");
> The unit test then does TEST_UTIL.createMultiRegions.
> You'll see all regions being created and then they UNASSIGNED znodes are created.  Grep
for the 'eee' row from testCachePrewarm table (be careful, there is also logging for testCachePrewarm2
in this log).
> The last thing you'll see is:
> 2010-06-19 19:34:39,628 DEBUG [RegionManager.metaScanner] master.RegionManager(1006):
Created UNASSIGNED zNode testCachePrewarm,eee,1276976076048.557068905bf2abfe84a5e49953a23c02.
in state M2ZK_REGION_OFFLINE
> For ddd and fff, you'll see those assigned out.
> Maybe the test is not hanging about long enough but wonder why its skipped?

-- 
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