[pkg-lxc-devel] Bug#847048: Bug#847048: lxc: centos6 fails to start after u/g to 2.0.6

[Dean Hamstead]

Yes, Solution was to use "vsyscall=emulate" on the host kernel command 
line (ie on my laptop).

Exactly the same issue as: 
https://github.com/CentOS/sig-cloud-instance-images/issues/62

So I am guessing the issue was actually the kernel upgrading.


See also: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1018415




On 05/12/16 21:42, Evgeni Golov wrote:
> On Mon, Dec 05, 2016 at 08:59:07PM +1100, Dean Hamstead wrote:
>> it seems if i drop back to 2.0.5 the problem still occurs. so i am guessing
>> some other thing upgraded and its not lxc-start's fault
> can you try an older lxcfs, it was updated at (almost) the same time.
> mind you, just downgrading the package is not enough, the daemon is not
> restarted automatically as this would break running containers. restart
> it yourself or reboot :)
>
>> On 05/12/16 19:25, Evgeni Golov wrote:
>>> Hi Dean,
>>>
>>> thanks for the report,
>>>
>>> On Mon, Dec 05, 2016 at 05:58:40PM +1100, Dean Hamstead wrote:
>>>> Since upgrading, centos6 containers fail to start.
>>> Is that a regression from 2.0.5? Or anything older?
>>>
>>>> I am guessing that upstart is doing something crappy and lxc is now carefull enough to notice?
>>> Could totally be, but I would guess upstream would have tested with
>>> upstart (on older Ubuntu LTS releases). Well, let's see if I can
>>> reproduce tonight.
>>>
>>> Greets
>>> Evgeni