[
https://issues.apache.org/jira/browse/TAJO-1440?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14374990#comment-14374990
]
Dongjoon Hyun commented on TAJO-1440:
-------------------------------------
To [~jihoonson] and [~eminency], after rebooting, I can not reproduce that
error.
{code}
mvn clean install -Pparallel-test,hcatalog-0.12.0 -DLOG_LEVEL=ERROR
-Dmaven.fork.count=2 > TAJO-1440.travis.log6.ERROR.txt
mvn clean install -Pparallel-test,hcatalog-0.12.0 -DLOG_LEVEL=ERROR
-Dmaven.fork.count=2 > TAJO-1440.travis.log7.ERROR.txt
mvn clean install -Pparallel-test,hcatalog-0.12.0 -DLOG_LEVEL=ERROR
-Dmaven.fork.count=2 > TAJO-1440.travis.log8.ERROR.txt
mvn clean install -Pparallel-test,hcatalog-0.12.0 -DLOG_LEVEL=ERROR
-Dmaven.fork.count=2 > TAJO-1440.travis.log9.ERROR.txt
mvn clean install -Pparallel-test,hcatalog-0.12.0 -DLOG_LEVEL=ERROR
-Dmaven.fork.count=2 > TAJO-1440.travis.log10.ERROR.txt
{code}
The results are all successful.
https://app.box.com/s/qv845vtle466ie3zz6fn4lcd2d16wz4p
https://app.box.com/s/k9akpw1h245w73db9zh7kakxw0ebkrza
https://app.box.com/s/qdu77lqcwexbx9wzrw9tdf01yf4rhlnf
https://app.box.com/s/gneixm5umubl1hn0xm0ri3c7juwbj53x
https://app.box.com/s/w11vkajav48evhy8atdxdth9kuz6nx4m
I'm very sorry for this. In daytime, I met that error everytime I built if I
use ERROR level.
By the way, I should change my branch into TAJO-1403 'SimpleQuery', so I
stopped today's testing here.
> Some tests fail in parallel test environment
> --------------------------------------------
>
> Key: TAJO-1440
> URL: https://issues.apache.org/jira/browse/TAJO-1440
> Project: Tajo
> Issue Type: Bug
> Components: test
> Affects Versions: 0.10.0
> Reporter: Jongyoung Park
> Assignee: Jongyoung Park
> Priority: Blocker
>
> Running following command, test process often fails.
> bq. mvn clean install -Pparallel-test,hcatalog-0.12.0 -DLOG_LEVEL=ERROR
> -Dmaven.fork.count=2
> But it is not 100% failure rate.
> It looks like a concurrency problem.
> The reason why is :
> # When INFO instead of ERROR for log level is used, it doesn't seem that it
> happens.
> # Only it happens in parallel test.
> Usual failure spot is at TestKillQuery.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
