On Fri, Mar 08, 2019 at 02:08:39PM +0000, Mark Hindley wrote:
> On Fri, Mar 08, 2019 at 03:00:31PM +0100, KatolaZ wrote:
> > > Yes, elogind uses it. But I got upstream to take a patch which uses
> > > /var/lib/dbus/machine-id as an alternative to /etc/machine-id.
> > >
> >
> > and, IIRC, also /var/lib/dbus/machine-id is re-generated at boot
> > time. But we need to double-check.
>
> Yes, but only if it is missing, so eg if /var/ is tmpfs or you arrange to clear
> it.
>
> /var/lib/dbus/machine-id is generated by a dbus-uuidgen(1) call in the dbus
> initscript. This doesn't overwrite an existing file so that the machine-id is
> not changed during a single uptime.
>
> Certainly elogind has no requirement for /etc/machine-id to be present or
> consistent between boots.

Then we could probably just ignore it, right? It turns out I was
making confusion between /etc/machine-id and
/car/lib/dbus/machine-id. There is no /etc/machine-id in any of my
machines, either on ascii or on beowulf, and I had forgotten you had
pushed the patch to elogind to use the dbus machine-id instead.

We should just double-check that the dbus machine-id is regenerated at
reboot, as expected.

HND

KatolaZ

-- 
[ ~.,_  Enzo Nicosia aka KatolaZ - Devuan -- Freaknet Medialab  ]  
[     "+.  katolaz [at] freaknet.org --- katolaz [at] yahoo.it  ]
[       @)   http://kalos.mine.nu ---  Devuan GNU + Linux User  ]
[     @@)  http://maths.qmul.ac.uk/~vnicosia --  GPG: 0B5F062F  ] 
[ (@@@)  Twitter: @KatolaZ - skype: katolaz -- github: KatolaZ  ]