[ 
https://issues.apache.org/jira/browse/MAPREDUCE-4980?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16129582#comment-16129582
 ] 

Allen Wittenauer commented on MAPREDUCE-4980:
---------------------------------------------

bq. I wasn't able to reproduce the TestNNBench failure locally

It's extremely racy and seems to only show up under heavy load. Fun!  You might 
try increasing the thread count.

bq. Looks like lots of things want to scribble on the same areas in 
target/test-dir.

The test-dir/x/dfs thing is a red herring.  The test dirs get re-used by 
different threads at different times by assigning the various system properties 
when they are launched.  You'll see the same sort of thing when running 
parallel in hadoop-hdfs; it's a product of the MiniDFSCluster.  It's a little 
confusing and I sort of wish MiniDFSCluster always unique'd the directory (but 
in a smarter way than MiniYARNCluster).  On the flip side, I'm too lazy to set 
the property that gives each test it's own HDFS base dir. ;)

I'd need to see stack traces or something else to help diagnose further. If you 
see any "x didn't cleanup in 20 secs" or whatever, those are the timing issues 
I'm still working out. Finding a not-insane multiplier that works on Jenkins 
reliably has been hard. Suspect there are a lot of dead processes on those 
machines. :(

BTW, there are a TON of expected and/or "useless" stack traces (I'm looking at 
you, metrics.) in the test output.  The logging doesn't provide a lot of hints 
and so be careful about that too.

> Parallel test execution of hadoop-mapreduce-client-core
> -------------------------------------------------------
>
>                 Key: MAPREDUCE-4980
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-4980
>             Project: Hadoop Map/Reduce
>          Issue Type: Test
>          Components: test
>    Affects Versions: 3.0.0-alpha1
>            Reporter: Tsuyoshi Ozawa
>            Assignee: Andrey Klochkov
>         Attachments: MAPREDUCE-4980.010.patch, MAPREDUCE-4980.011.patch, 
> MAPREDUCE-4980.012.patch, MAPREDUCE-4980.1.patch, MAPREDUCE-4980--n3.patch, 
> MAPREDUCE-4980--n4.patch, MAPREDUCE-4980--n5.patch, MAPREDUCE-4980--n6.patch, 
> MAPREDUCE-4980--n7.patch, MAPREDUCE-4980--n7.patch, MAPREDUCE-4980--n8.patch, 
> MAPREDUCE-4980.patch
>
>
> The maven surefire plugin supports parallel testing feature. By using it, the 
> tests can be run more faster.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

---------------------------------------------------------------------
To unsubscribe, e-mail: mapreduce-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: mapreduce-issues-h...@hadoop.apache.org

Reply via email to