[libvirt PATCH 09/15] systemd: check org.freedesktop.machine1 registration

[Daniel P. Berrangé]

On Tue, Apr 07, 2020 at 12:29:29PM +0200, Marc-André Lureau wrote:
> Hi
> 
> On Tue, Apr 7, 2020 at 11:50 AM Ján Tomko <jtomko at redhat.com> wrote:
> >
> > On a Monday in 2020, marcandre.lureau at redhat.com wrote:
> > >From: Marc-André Lureau <marcandre.lureau at redhat.com>
> > >
> > >Since commit f10bd740e178c89f24d0b0298d0b5413537d0699 ("Cache the
> > >presence of machine1 service"), the code checks for systemd1
> > >registration. Not totally unreasonable, but that seems odd since we
> > >actually check machined presence in this function.
> >
> > That is intentional.
> >
> > We only count the systemd-based services as really activatable
> > if systemd1 is already registered.
> >
> > On some Frankenstein'd Gentoo systems with systemd installed
> > but not started, the services were showing up as activatable
> > but failed with obscure errors.
> >
> > Some history:
> > commit 12ee0b98d3ddcb2c71cdb2352512153791ebcb7e
> >      Check if systemd is running before creating machines
> > https://libvirt.org/git/?p=libvirt.git;a=commitdiff;h=12ee0b98
> > https://bugs.gentoo.org/show_bug.cgi?id=493246#c22
> 
> Ok, I imagined something like that. However, making systemd a
> hard-dependency for machine1 is not a good solution. In theory,
> machine1 could have a different implementation that doesn't rely on
> systemd.

I don't think we need to worry about that possibilty, since AFAIK
no such alternate impl exists. Such a replacement would be a
major piece of work needing to implement a lot of system logic
around the cgroups/service setup, so I'd be sceptical that libvirt
would "just work" regardless. 

Regards,
Daniel
-- 
|: https://berrange.com      -o-    https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org         -o-            https://fstop138.berrange.com :|
|: https://entangle-photo.org    -o-    https://www.instagram.com/dberrange :|