[Users] "reboot" inside a container stops the container

[Kir Kolyshkin]

On 13 April 2011 13:16, Benjamin Henrion <bh at udev.org> wrote:
> Hi,
>
> I found out that typing "reboot" in a container halts the whole
> container, making it unreacheable (the container itself is not running
> anymore, checking with vzlist on the physical machine).

Which vzctl and kernel do you use?

In case of vzctl < 3.0.25 reboot is performed by the vpsreboot cron
script which is
called from cron every 5 minutes by default. Check if your cron daemon
is running,
check that you have that job installed, and note that since this job
is run every 5
minutes, your container can be done for as long as 5 minutes.

In case of vzctl >= 3.0.25 (and a decent/recent kernel) reboot is performed by a
daemon called vzeventd. Check that vzevent kernel module is loaded,
that vzeventd
daemon is running and included in autostart.