[lxc-users] lxc stop does not stop a CentOS 7 container

[Hans Deragon]

I found a bug report that is probably related to our issues:

"Shutdown failure: Assertion 'sd_id128_randomize(&id) >= 0' failed at 
../src/core/dbus.c:657, function bus_on_connection(). Aborting."
https://bugs.launchpad.net/ubuntu/+source/systemd/+bug/1519499

I am willing to help out with debugging, if anybody has time to work on 
this.

Best regards,
Hans Deragon

On 2016-05-09 09:37, Hans Deragon wrote:

> Greetings,
> 
> I have a similar problem with Xenial, though I have not found any 
> solution (I am not familiar with systemd and the solution provided for 
> CentOS does not apply to Xenial).
> 
> I created a few days ago a new container on an updated Ubuntu Trusty 
> Thar 14.04 LTS host using the following command:
> 
>  lxc launch images:ubuntu/xenial/amd64 server1
> 
> Now, it hangs when I run:
> 
>  lxc stop server1
> 
> This problem was reported at https://github.com/lxc/lxd/issues/1570 [3]
> 
> What puzzles me, is that stopping a container is such a basic feature 
> that I am surprise to find that it is not being addressed. Both Trusty 
> and Xenial are now out and production ready, yet one cannot stop a 
> Xenial container. Are we the only one that have this problem? Should I 
> open a new bug report or can someone open this one again since it is 
> not fixed?
> 
> Best regards,
> Hans Deragon
> 
> On 2016-05-06 13:14, Hiroaki Nakamura wrote:
> 
> > Hi all,
> >
> > I was trying lxc on Ubuntu 16.04 LTS and found out that
> > lxc stop does not stop a CentOS 7 container.
> >
> > I ran the following commands.
> >
> > lxc image copy images:/centos/7/amd64 local: --alias=cent7
> > lxc launch cent7 cent01
> > lxc stop cent01
> > ^ This does not stop the cent01 container even after one minute,
> > so I pressed Control-C to kill this command.
> >
> > I found a workaround at
> > https://lists.linuxcontainers.org/pipermail/lxc-users/2014-February/006304.html [1]
> >
> > Here is a command to fix.
> > lxc exec cent01 -- sh -c 'ln -s /usr/lib/systemd/system/halt.target
> > /etc/systemd/system/sigpwr.target && systemctl daemon-reload'
> >
> > After this fix, the container is stopped immediately with
> > lxc stop cent01
> >
> > I would like to this fix to be included in the original image.
> > How can I contact the maintainers of container images?
> >
> > Here are versions of lxc and CentOS 7 image.
> >
> > $ lxc version
> > 2.0.0
> >
> > $ lxc image list cent
> > +-------+--------------+--------+-----------------------------------+--------+---------+-----------------------------+
> > | ALIAS | FINGERPRINT | PUBLIC | DESCRIPTION |
> > ARCH | SIZE | UPLOAD DATE |
> > +-------+--------------+--------+-----------------------------------+--------+---------+-----------------------------+
> > | cent7 | a027d59858d6 | no | Centos 7 (amd64) (20160506_02:16) |
> > x86_64 | 62.88MB | May 6, 2016 at 9:55am (UTC) |
> > +-------+--------------+--------+-----------------------------------+--------+---------+-----------------------------+
> >
> > Regards,
> > Hiroaki