You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@mesos.apache.org by "Anand Mazumdar (JIRA)" <ji...@apache.org> on 2017/04/12 17:18:42 UTC

[jira] [Commented] (MESOS-7381) Flaky tests in NestedMesosContainerizerTest

    [ https://issues.apache.org/jira/browse/MESOS-7381?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15966238#comment-15966238 ] 

Anand Mazumdar commented on MESOS-7381:
---------------------------------------

[~gilbert] [~jieyu] Looks severe enough to be fixed for 1.3.0. Can you guys help triage this?

> Flaky tests in NestedMesosContainerizerTest
> -------------------------------------------
>
>                 Key: MESOS-7381
>                 URL: https://issues.apache.org/jira/browse/MESOS-7381
>             Project: Mesos
>          Issue Type: Bug
>          Components: test
>    Affects Versions: 1.1.1, 1.2.0
>            Reporter: Zhitao Li
>            Priority: Critical
>
> See https://gist.github.com/zhitaoli/40aacc3f2ecb88802ee58f40c46d43f9 for a log.
> There are about 12-13 tests in that class which are very flaky in my environment (Debian/jessie running 4.4.38 kernel).
> It seems like the primary cause is that root container "failed" and caused `reap` to be called on itself, which cascalade and cause both containers to be actively killed by containerizer instead of terminate by themselves.
> This happens on master branch at commit 6c1e20c0f2777d9bb831be3ff43c885b253af7bb



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)