Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

NO LOGINS: System going down in 30 seconds???

732 views
Skip to first unread message

Neill Griffin

unread,
Jun 19, 1998, 3:00:00 AM6/19/98
to

Howdy,
a really dumb question for you.

I've just restarted a SS1000/Solaris 2.5.1 after a nice orderly shutdown
(or so I thought) and now when I try to telnet into it is accepts the
login name and then immediately sez "NO LOGINS: System going down in 30
seconds " and closes the connection.

The console is working fine, as is the Web server and other processes,
so is there something that hasn't been cleaned up after the restart?

Puzzled

TIA

ng
mrb...@mira.net

Neil Hoggarth

unread,
Jun 19, 1998, 3:00:00 AM6/19/98
to

In article <358A3A9B.4330@look_in_sig.net>,
Neill Griffin <mrbean@look_in_sig.net> wrote:

> I've just restarted a SS1000/Solaris 2.5.1 after a nice orderly shutdown
> (or so I thought) and now when I try to telnet into it is accepts the
> login name and then immediately sez "NO LOGINS: System going down in 30
> seconds " and closes the connection.

Remove the file "/etc/nologin". This is written by the shutdown
process to prevent users coming onto the system just as it is about to
go down. I think that it is normally removed by the system as it boots
- I don't know why you're stuck with it.

Regards,
--
+-------------------------------------------------------------------------+
Neil Hoggarth Departmental Computer Officer
<neil.h...@physiol.ox.ac.uk> Laboratory of Physiology
http://www.physiol.ox.ac.uk/~njh/ Oxford University, UK
+-------------------------------------------------------------------------+

Steve Bellenot

unread,
Jun 19, 1998, 3:00:00 AM6/19/98
to

In article <358A3A9B.4330@look_in_sig.net>,
Neill Griffin <mrbean@look_in_sig.net> wrote:
>Howdy,
> a really dumb question for you.
>
>I've just restarted a SS1000/Solaris 2.5.1 after a nice orderly shutdown
>(or so I thought) and now when I try to telnet into it is accepts the
>login name and then immediately sez "NO LOGINS: System going down in 30
>seconds " and closes the connection.
>
>The console is working fine, as is the Web server and other processes,
>so is there something that hasn't been cleaned up after the restart?
>
>Puzzled
>
>TIA
>
>ng
>mrb...@mira.net

The way it was in 4.1.x
/etc/nologin (or something similar) was created by the shutdown command
and login would not let anyone remain logined in when it existed. [It
would just cat the the file /etc/nologin and quit. The boot sequence
would remove /etc/nologin.

I believe this behavoir is not present in 5.x by default, but it is
easy to add via the global .profile .cshrc system files in /etc [with
different names].

If true, you might be able to telnet in as root, and remove the file.
Or boot single user, and remove it. Or boot from CD and remove it.
--
bell...@math.fsu.edu http://www.math.fsu.edu/~bellenot (850)644-7189 (4053fax)


Mike J Tietel

unread,
Jun 19, 1998, 3:00:00 AM6/19/98
to

In article <6me1np$8jh$1...@news.fsu.edu>, bell...@math.fsu.edu (Steve Bellenot) writes:
> The way it was in 4.1.x
> /etc/nologin (or something similar) was created by the shutdown command
> and login would not let anyone remain logined in when it existed. [It
> would just cat the the file /etc/nologin and quit. The boot sequence
> would remove /etc/nologin.
>
> I believe this behavoir is not present in 5.x by default, but it is
> easy to add via the global .profile .cshrc system files in /etc [with
> different names].
>
> If true, you might be able to telnet in as root, and remove the file.
> Or boot single user, and remove it. Or boot from CD and remove it.
>

The /etc/nologin functionality did not exist in Solaris 2.0 - 2.3, you
had to "roll your own". I don't remember if 2.4 had it or not, but it
is back as of 2.5...

You probably cannot telnet in - telnet requires login...

You can FTP in and remove /etc/nologin, however.

--
Mike Tietel Consultant - Euler Solutions, Inc.
Currently at:
mike_...@adc.com ADC Telecommunications, Inc
(612) 946-2272 P.O. Box 1101, MS-254
(612) 946-3910 (fax) Minneapolis, MN 55440-1101

Neill Griffin

unread,
Jun 20, 1998, 3:00:00 AM6/20/98
to

Steve Bellenot wrote:
>
> In article <358A3A9B.4330@look_in_sig.net>,
> Neill Griffin <mrbean@look_in_sig.net> wrote:
> >Howdy,
> > a really dumb question for you.
> >
> >I've just restarted a SS1000/Solaris 2.5.1 after a nice orderly shutdown
> >(or so I thought) and now when I try to telnet into it is accepts the
> >login name and then immediately sez "NO LOGINS: System going down in 30
> >seconds " and closes the connection.
> >
> >The console is working fine, as is the Web server and other processes,
> >so is there something that hasn't been cleaned up after the restart?
> >
> >Puzzled
> >
> >TIA
> >
> >ng
> >mrb...@mira.net
>
> The way it was in 4.1.x
> /etc/nologin (or something similar) was created by the shutdown command
> and login would not let anyone remain logined in when it existed. [It
> would just cat the the file /etc/nologin and quit. The boot sequence
> would remove /etc/nologin.
>
> I believe this behavoir is not present in 5.x by default, but it is
> easy to add via the global .profile .cshrc system files in /etc [with
> different names].
>
> If true, you might be able to telnet in as root, and remove the file.
> Or boot single user, and remove it. Or boot from CD and remove it.
> --
Hi,
thanks for the prompt responses. One thing I did check was for any
new files in /etc as this was the most likely place for something like
this, but I didn't notice this file... will check again. Yes I do have
console and ftp (root) access, so I will investigate further.

Thanks again

ng

Neill Griffin

unread,
Jun 20, 1998, 3:00:00 AM6/20/98
to

Howdee,
well as everyone had correctly pointed out, there WAS a
/etc/nologin file in existance which was stopping any telnet connections
after the restart. I was using WS-FTP 5 to view the directory and I
perhaps had forgotton to refresh the display... or I just plain missed
it... it's pretty obvious but :)

The next question was, why wasn't it 'cleaned' up after the restart?

The answer turns out to be that someone??? had updated the shutdown
script across our two 2.5.1 servers to the same version, but the problem
one was running an earlier version of the /etc/rc2.d/S05RMTMPFILES
(v.1.13 vs v1.16) script which did not clean up the nologin file. I'm
not sure when this was done as the timestamp on the shutdown scripts was
01/01/70... duh! or why they weren't running the same release of scripts
across the board.

Thanks everyone

ng
mrb...@mira.net

0 new messages