Stale /var/run/nologin

[Sam Varshavchik]

Steven P. Ulrick writes:

> On Sat, 4 Jan 2014 19:42:16 -0600
> Steven Ulrick <meow8282 at gmail.com> wrote:
>
> > On Saturday, January 4, 2014, Sam Varshavchik <mrsam at courier-mta.com>
> > wrote:
> > > About ten hours after a reboot, a chance attempt to log in back to
> > > the
> > server was rather rudely rejected with a:
> > >
> > > System is booting up. See pam_nologin(8)
> > > Connection closed by 192.168.0.2
> > >
> > > A quick run back to the console revealed the existence of a ten
> > > hour old
> > /var/run/nologin file as the culprit. Removing it put everything back
> > in working order.
> >
> > Two things:
> > 1. You are not alone!
> > 2. Thanks for the workaround...
>
> OK...  I guess #2 was a bit premature.  I just had reason to log out of
> my KDE session.  When I logged back in (or tried to), I was greeted
> with the dreaded "System is booting up. See pam_nologin(8)"  When my
> system is in this state, I can get to a console, but it will not let me
> log in.  I even tried to ssh from another system.  The other system
> informs me of the following:
> > > System is booting up. See pam_nologin(8)
>
> For clarification, the remote system that I am attempting to ssh FROM
> is telling me that the system that I am trying to ssh INTO is in the
> following state:
> > > System is booting up. See pam_nologin(8)
>
> So, it appears that I have no workaround to this issue other than
> rebooting...

No. According to the documentation, root is allowed to log in. You should be  
able to ssh as root.

My original working theory was that /run/nologin was not getting cleared by  
whatever godforsaken systemd service is responsible for removing it, when  
the boot mostly completes. This was based on my /run/nologin's timestamp,  
which dated back to my system's boot.

But if that's getting spuriously created, during a normal system state, then  
something indeed must be creating it, in the wild.

Not going to be easy tracking it down. Perusing journalctl's man page, there  
doesn't seem to be a way to specify a time interval. Given /run/nologin's  
timestamp, it should be possible to track down what was started in that  
timeframe, but I do not see a way to specify a timeframe. Furthermore,  
journalctl's output seems to consist of merely log messages from systemd- 
started processes, rather than the actual log of what was started, and when.

So, tell me again how logs kept as binary blobs are superior to plain text  
files.

I'd start to hit Google, looking for way to find systemd's actual logs, and  
filtering them by a time interval. Seems silly to have to do that, but  
systemd is such a winner…

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 819 bytes
Desc: not available
URL: <http://lists.fedoraproject.org/pipermail/users/attachments/20140108/75a32393/attachment.sig>