Re: [PATCH container 1/3] setup: add no-op check_systemd_nesting for unmanaged CTs

[Robert Obkircher <r.obkircher@proxmox.com>]

On 2/4/26 10:16, Daniel Kral wrote:
> Otherwise the container will fail to start, create and clone, because
> the plugin's check_systemd_nesting is only defined in the Base module
> but not the Unmanaged module and is called in the pre_start_hook,
> post_clone_hook, and post_create_hook.
>
> Reported in the Proxmox Forum [0].
>
> [0] https://forum.proxmox.com/threads/180258/
>
> Signed-off-by: Daniel Kral <d.kral@proxmox.com>
> ---
>  src/PVE/LXC/Setup/Unmanaged.pm | 8 +++++++-
>  1 file changed, 7 insertions(+), 1 deletion(-)
>
> diff --git a/src/PVE/LXC/Setup/Unmanaged.pm b/src/PVE/LXC/Setup/Unmanaged.pm
> index d76921e..aa26c1c 100644
> --- a/src/PVE/LXC/Setup/Unmanaged.pm
> +++ b/src/PVE/LXC/Setup/Unmanaged.pm
> @@ -51,7 +51,13 @@ sub unified_cgroupv2_support {
>  
>  sub get_ct_init_path {
>      my ($self) = @_;
> -    return '/sbin/init'; # only passed to unified_cgroupv2_support for now
> +    # only passed to check_systemd_nesting and unified_cgroupv2_support for now
> +    return '/sbin/init';
> +}
> +
> +sub check_systemd_nesting {
> +    my ($self, $conf, $init) = @_;
> +    return;
>  }

Applying this patch fixed the startup problem. I tested this by
changing the ostype of a Debian container to unmanaged.

The only thing I've noticed is that the method is also missing from
Plugin.pm.

>  
>  sub ssh_host_key_types_to_generate {