[
https://issues.apache.org/jira/browse/HBASE-10410?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13880484#comment-13880484
]
Andrew Purtell commented on HBASE-10410:
----------------------------------------
Or this could be https://jira.codehaus.org/browse/SUREFIRE-817, and I think no
test failed because nothing was printed after timeout or some other unexpected
exit. I notice we are still running with our custom Surefire version
2.12-TRUNK-HBASE-2. The latest Surefire plugin release is 2.16. Any reason not
to be using that one? Ping [~stack], [~nkeywal].
Trying it locally. One difference is ResourceChecker INFO logs are now showing
up on the console.
> On some JREs, hbase-server is failing due to "timeout" without any test
> actually timing out
> -------------------------------------------------------------------------------------------
>
> Key: HBASE-10410
> URL: https://issues.apache.org/jira/browse/HBASE-10410
> Project: HBase
> Issue Type: Bug
> Affects Versions: 0.98.0
> Reporter: Andrew Purtell
>
> This is frustrating. On some JREs (Java 7) I am seeing hbase-server fail due
> to timeout, without any test actually reporting a timeout. This
> StackOverflow answer talks about how JUnit < 4.12-SNAPSHOT may report a
> timeout if the test throws an InterruptedException:
> http://stackoverflow.com/questions/17016011/junit-test-times-out-despite-executing-quickly
> Those happen at various points when shutting down the minicluster. We may
> be letting one escape.
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
