[
https://issues.apache.org/jira/browse/HBASE-4690?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13139543#comment-13139543
]
Ted Yu commented on HBASE-4690:
-------------------------------
It is pretty clear what happened in build 2384. The failure was because regions
brought online wasn't in the same order as start keys are defined:
{code}
2011-10-29 21:45:43,789 INFO
[RS_OPEN_REGION-hemera.apache.org,37045,1319924731557-0]
regionserver.HRegion(502): Onlined
observed_table,kkk,1319924743536.d2bb03652b0e69a4a192be3b60f6cd78.; next
sequenceid=1
...
2011-10-29 21:45:43,883 DEBUG [Thread-183] hbase.HBaseTestingUtility(1129):
Found 25 rows for table observed_table
2011-10-29 21:45:43,883 DEBUG [Thread-183] hbase.HBaseTestingUtility(1132):
FirstRow=observed_table,,1319924743504.ed6d9b9f5122809fad16e61835367b48.
2011-10-29 21:45:43,887 INFO
[RS_OPEN_REGION-hemera.apache.org,37045,1319924731557-1]
regionserver.HRegion(502): Onlined
observed_table,lll,1319924743540.ac163536355dbe1ab71ab1a9ee7a22d4.; next
sequenceid=1
...
2011-10-29 21:45:43,950 DEBUG [main-EventThread] zookeeper.ZKUtil(228):
master:34047-0x13351a50a270000 Set watcher on existing znode
/hbase/unassigned/ed6d9b9f5122809fad16e61835367b48
...
2011-10-29 21:45:44,050 INFO
[RS_OPEN_REGION-hemera.apache.org,45759,1319924731527-0]
regionserver.HRegion(502): Onlined
observed_table,,1319924743504.ed6d9b9f5122809fad16e61835367b48.; next
sequenceid=1
{code}
We can see the ~170ms delay between the discovery of region
1319924743504.ed6d9b9f5122809fad16e61835367b48. and its actual online.
A simple patch would be to give getRSForFirstRegionInTable() some time if index
returned by hbaseCluster.getServerWith() was -1.
> Intermittent
> TestRegionServerCoprocessorExceptionWithAbort#testExceptionFromCoprocessorDuringPut
> failure
> --------------------------------------------------------------------------------------------------------
>
> Key: HBASE-4690
> URL: https://issues.apache.org/jira/browse/HBASE-4690
> Project: HBase
> Issue Type: Test
> Affects Versions: 0.92.0
> Reporter: Ted Yu
> Assignee: Eugene Koontz
> Fix For: 0.92.0
>
>
> See
> https://builds.apache.org/view/G-L/view/HBase/job/HBase-0.92/83/testReport/junit/org.apache.hadoop.hbase.coprocessor/TestRegionServerCoprocessorExceptionWithAbort/testExceptionFromCoprocessorDuringPut/
> Somehow getRSForFirstRegionInTable() wasn't able to retrieve the region
> server.
> One fix for this issue is to spin up MiniCluster with 1 region server so that
> we don't need to search for the region server where first region is hosted.
--
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
