[jira] [Commented] (HBASE-8069) TestHLog is dependent on the execution order

["stack (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/HBASE-8069?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13599387#comment-13599387 ] 

stack commented on HBASE-8069:
------------------------------

+1 on commit.
                
> TestHLog is dependent on the execution order
> --------------------------------------------
>
>                 Key: HBASE-8069
>                 URL: https://issues.apache.org/jira/browse/HBASE-8069
>             Project: HBase
>          Issue Type: Bug
>          Components: test, wal
>    Affects Versions: 0.95.0, 0.94.5
>            Reporter: Matteo Bertozzi
>            Assignee: Matteo Bertozzi
>         Attachments: HBASE-8069-v0.patch, HBASE-8069-v0.patch
>
>
> Changing the execution order of the tests, TestHLog show up as failing
> {code}<runOrder>alphabetical</runOrder>{code}
> testAppendClose() changes the DFS cluster of the test (set globally) so the tests executed after has the new DFS. Trying to start a new mini cluster for every test @Before seems to solve the problem.
> {code}
> testSplit(org.apache.hadoop.hbase.regionserver.wal.TestHLog): 3 exceptions [org.apache.hadoop.ipc.RemoteException(org.apache.hadoop.hdfs.server.namenode.LeaseExpiredException): No lease on /user/th30z/hbase/TestHLog/21985ea121a1d65fa82d90d521be7313/recovered.edits/0000000000000000001.temp File is not open for writing. Holder DFSClient_NONMAPREDUCE_1044150136_583 does not have any open files.
> {code}
> Looking at testSplit() the failure seems to be in logSplitter.splitLog(), OutputSink.finishWritingAndClose() is not able to close the files and rethrow the exception.

--
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