[
https://issues.apache.org/jira/browse/HBASE-3100?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
stack updated HBASE-3100:
-------------------------
Attachment: 3100-part3.txt
Upping compaction threshold did not prevent splits. A split went ahead because
there were 'References'. Trying another config. Also, set an upper bound
timer of 5 minutes on this test. Turns out this test failed on hudson and it
cause TestTableMapReduce to fail (hudson #1541) because it left a zk running.
TTMR put up its own zk on a different port but last task in the MR job suddenly
finds this tests ZK instead of the TTMR one (looking into that too but
committing this in meantime).
> TestMergeTable failing in TRUNK
> -------------------------------
>
> Key: HBASE-3100
> URL: https://issues.apache.org/jira/browse/HBASE-3100
> Project: HBase
> Issue Type: Bug
> Reporter: stack
> Assignee: stack
> Fix For: 0.90.0
>
> Attachments: 3100-part3.txt, 3100-xtra-debug.txt, 3100.txt
>
>
> This is an old crufty test that is failing in TRUNK. We create regions --
> user, root, and meta regions offline -- but then come along and disable them
> after starting up minicluster. Confuses new master. This test also does not
> emit logging up in hudson because its written against old-style testing...
> even has a subclass of the HBaseTestCase to do this crazy manual creation.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
