You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@mesos.apache.org by "Benjamin Bannier (JIRA)" <ji...@apache.org> on 2017/01/11 16:13:58 UTC
[jira] [Commented] (MESOS-6910) Segfault when running agent on
anything other than Linux while using unified containerizer
[ https://issues.apache.org/jira/browse/MESOS-6910?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15818737#comment-15818737 ]
Benjamin Bannier commented on MESOS-6910:
-----------------------------------------
[~aaron.wood] This looks extremely similar to MESOS-5745, especially the last two frames. The issue there is UB in libc++ which clang started to optimize. Are you sure you are not using any of the affected clang or libc++ versions? Note that e.g., {{_assign_multi}} also is an implementation detail in libc++ while e.g., libstdc++ (as of {{cc1e50b13366d00f1e065985dac63a56055e3e08}}) does not contain any mention of this function.
> Segfault when running agent on anything other than Linux while using unified containerizer
> ------------------------------------------------------------------------------------------
>
> Key: MESOS-6910
> URL: https://issues.apache.org/jira/browse/MESOS-6910
> Project: Mesos
> Issue Type: Bug
> Components: agent, containerization, isolation
> Affects Versions: 1.1.0
> Reporter: Aaron Wood
>
> When running the Mesos agent on OS X ([~hausdorff] not sure if this applies to Windows or not) and using the unified containerizer the agent will segfault:
> WARNING: Logging before InitGoogleLogging() is written to STDERR
> I0110 12:45:56.882041 2100015104 main.cpp:243] Build: 2016-11-16 18:56:50 by brew
> I0110 12:45:56.883872 2100015104 main.cpp:244] Version: 1.1.0
> I0110 12:45:56.887733 2100015104 containerizer.cpp:200] Using isolation: docker/runtime,filesystem/posix
> Aborted at 1484070356 (unix time) try "date -d @1484070356" if you are using GNU date ***
> PC: @ 0x10b5049ca std::__1::__tree<>::__assign_multi<>()
> SIGSEGV (@0x0) received by PID 47147 (TID 0x7fff7d2bb000) stack trace: ***
> @ 0x7fff899b252a _sigtramp
> @ 0x7feea043f870 (unknown)
> @ 0x10b50496c flags::FlagsBase::operator=()
> @ 0x10b503924 mesos::uri::fetcher::Flags::operator=()
> @ 0x10b502902 mesos::uri::fetcher::create()
> @ 0x10b4eefd3 mesos::internal::slave::docker::Store::create()
> @ 0x10b4b241f mesos::internal::slave::Store::create()
> @ 0x10b49925a mesos::internal::slave::Provisioner::create()
> @ 0x10b40cc6f mesos::internal::slave::MesosContainerizer::create()
> @ 0x10b3ab426 mesos::internal::slave::Containerizer::create()
> @ 0x10af20d80 main
> @ 0x7fff9a3995ad start
> @ 0x6 (unknown)
> Segmentation fault: 11
> To reproduce this you can run:
> mesos-slave --master=127.0.0.1:5050 --work_dir=/tmp/slave --containerizers=mesos --image_providers=docker --isolation=docker/runtime
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)