[lxc-users] Systemd Crash

Fajar A. Nugraha list at fajar.net
Mon Aug 31 20:21:36 UTC 2015


On Tue, Sep 1, 2015 at 1:29 AM, Federico Alves <venefax at gmail.com> wrote:

> I get this crash every few minutes
> [619187.189766] systemd-coredump[1309752]: Failed to get COMM, falling
> back to the command line.
> [619187.189787] systemd-coredump[1309752]: Failed to get EXE.
>
> The host is FC 22 and the LXC-Container is FC 20
>


Fedora 20 reached EOL in June. Using that means you're on your own when it
comes to errors/bug fixes.


>
> At the container I have
> rpm -qa | grep systemd
> systemd-208-31.fc20.x86_64
> systemd-libs-208-31.fc20.x86_64
>
>
The earliest version of systemd that I tested and works good-enough for lxc
was 219, with 224 bringing even more improvement (e.g. only
systemd-journald fails on unprivileged container). Combined with the fact
that F20 is EOL, I doubt you'll get any help sticking with that
combination. The best option would probably to use F22 containers instead.

Another option, if your app ABSOLUTELY must run on F20 (e.g. due to
specific lib version requirements), AND you know how to rebuild your own
RPMs (with small fixes/modifications when necessary), AND you know how to
support your own installations (e.g. backporting security fixes yourself),
you could try backport latest systemd (225) from rawhide and try to get it
to work on F20.

-- 
Fajar
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linuxcontainers.org/pipermail/lxc-users/attachments/20150901/33c289a7/attachment.html>


More information about the lxc-users mailing list