Bug#769069: systemd: Failed to start Login Service

Paul Menzel pm.debian at googlemail.com
Mon Nov 17 22:26:48 GMT 2014


Am Montag, den 17.11.2014, 23:08 +0100 schrieb Sjoerd Simons:
> On Sat, Nov 15, 2014 at 01:16:36AM +0100, Paul Menzel wrote:
> > Am Freitag, den 14.11.2014, 10:18 +0000 schrieb Simon McVittie:
> > > > signal 15, shutting down normally. Nov 11 07:46:16 asrock-e350m1
> > > > avahi-daemon[929]: Got SIGTERM, quitting.
> > > ...
> > > > Nov 11 07:46:52 asrock-e350m1 NetworkManager[942]: <info> startup 
> > > > complete Nov 11 07:46:52 asrock-e350m1 NetworkManager[942]: <info>
> > > > exiting (success)
> > > 
> > > Any idea why these might have happened? They both use libdbus, so
> > > getting disconnected from D-Bus with exit-on-disconnect enabled would
> > > cause an _exit(1) rather than the raise(SIGTERM) you'd get from GDBus.
> > > 
> > > It looks vaguely as though *everything* on the system bus might have
> > > been failing to connect to it?
> > > 
> > > Did you / do you have any unusual systemd or dbus configuration in use?
> > 
> > Not that I know of. I am using the packaged configuration.
> 
> Yes this indeed looks like nothing can connect to the systemd bus.
>  
> > > If you can't reproduce this, then I don't think we'll be able to work
> > > out what is going on from that log, sorry.
> > 
> > It happened again, but I forgot to add the systemd debug stuff. It
> > stopped for a while after the start message of D-Bus and then continued
> > until the point where the “Logind failed to start” messages were shown.
> 
> One thing stood out:
> > Nov 14 23:39:05 asrock-e350m1 systemd[1]: var.mount: Directory /var to mount over is not empty, mounting anyway.
> 
> The dbus system socket is in unix:path=/var/run/dbus/system_bus_socket. So one
> possible explanation might be that your just mounted a file system over the
> /var/run dbus got its socket put in..
> 
> In principle this should not occur, but lets first work out if my theory is
> correct. Could you:
> 
> 0: Check where your final /var/run/ is pointing to (should a symlink to /run)

From a working boot, the symlink is correct.

        $ ls -l /var/run
        lrwxrwxrwx 1 root root 4 Mai 17  2011 /var/run -> /run

> 1: If it isn't a symlink, can you check whats in both the /var/run of your /var
>    partition *and* the var of your / partition (to see if you masked out the
>    dbus socket in any way)

When I hit this bug, I could not log in (probably due to logind not
running). The login prompt was shown and I was able to enter a password
but I could not log in and the shell was not usable.

Can I check what is going wrong in the systemd debug shell next time I
hit the problem?


Thanks,

Paul
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 181 bytes
Desc: This is a digitally signed message part
URL: <http://alioth-lists.debian.net/pipermail/pkg-systemd-maintainers/attachments/20141117/8212cbd5/attachment-0002.sig>


More information about the Pkg-systemd-maintainers mailing list