4:15 a.m.
Hi list,
We have a multi-workstation network, with several nodes running F16/XFCE
on remote-mounted home directories. All users can log in and out without
any difficulty except one. This user gives name and password but then
freezes indefinitely. 'ps' shows a variety of tasks running, but nothing
appears on screen except the standard wallpaper; time is present but
nothing else. Running are startxfce, xfce-session, etc. User can be
cleared by "kill -15 <startxfce-pid> and other users can log then log in.
What might be causing the freeze - is there some lockfile that needs
removing to unlock the user's account?
Jonathan
7:13 p.m.
Since it has been two days with no response, and this kind of reminds
me of a problem I had several months back (but who knows if it is
related, thread at
<http://lists.fedoraproject.org/pipermail/users/2011-December/thread.html#...>),
On Fri, Apr 20, 2012 at 6:15 PM, Jonathan Allen
<jonathan(a)barumtrading.co.uk> wrote:
Hi list,
We have a multi-workstation network, with several nodes running F16/XFCE
on remote-mounted home directories.
Is the remote mounted home directory for the affected user accessible?
All users can log in and out without
any difficulty except one. This user gives name and password but then
freezes indefinitely.
Can the user log in non-X11 via ssh or on a virtual console? How about
logging in from a different box?
'ps' shows a variety of tasks running, but nothing
appears on screen except the standard wallpaper; time is present but
nothing else. Running are startxfce, xfce-session, etc.
How about xfwm4?
User can be
cleared by "kill -15 <startxfce-pid> and other users can log then log in.
Are other users blocked on the one box or all over the network?
What might be causing the freeze - is there some lockfile that needs
removing to unlock the user's account?
Jonathan
Have you tried the brutal method of deleting the caches or the xfce state?
--
Joel Rees
1:51 a.m.
Joel,
Since it has been two days with no response, and this kind of
reminds
me of a problem I had several months back ...
> We have a multi-workstation network, with several nodes running F16/XFCE
> on remote-mounted home directories.
Is the remote mounted home directory for the affected user accessible?
Yes, quite happily on all the machines around the network.
> All users can log in and out without any difficulty except one.
This
> user gives name and password but then freezes indefinitely.
Can the user log in non-X11 via ssh or on a virtual console? How about
logging in from a different box?
Yes, a text console login and ssh from another machine both work fine.
> 'ps' shows a variety of tasks running, but nothing
> appears on screen except the standard wallpaper; time is present but
> nothing else. Running are startxfce, xfce-session, etc.
How about xfwm4?
Yes, that appears to be running, state 'S'
> User can be cleared by "kill -15 <startxfce-pid> and
other users can
> then log in.
Are other users blocked on the one box or all over the network?
No-one else can get a graphical session on that box because the graphical
console is taken with the stalled session, but console logins are still
fine and other users can freely start graphic sessions on other machines
on the network.
Have you tried the brutal method of deleting the caches or the xfce
state?
No - how do I do that?
Jonathan
2:03 a.m.
On 04/22/2012 11:51 PM, Jonathan Allen wrote:
No-one else can get a graphical session on that box because the
graphical
console is taken with the stalled session, but console logins are still
fine and other users can freely start graphic sessions on other machines
on the network.
What happens if somebody else tries to log on just after a reboot?
2:47 a.m.
Joe,
>No-one else can get a graphical session on that box because the
graphical
>console is taken with the stalled session, but console logins are still
>fine and other users can freely start graphic sessions on other machines
>on the network.
What happens if somebody else tries to log on just after a reboot?
That too is fine; any other user, either after a reboot or after the
frozen XFCE session has been taken out by "kill -15"
Jonathan
7:30 p.m.
On Mon, Apr 23, 2012 at 4:47 PM, Jonathan Allen
<jonathan(a)barumtrading.co.uk> wrote:
Joe,
> >No-one else can get a graphical session on that box because the graphical
> >console is taken with the stalled session, but console logins are still
> >fine and other users can freely start graphic sessions on other machines
> >on the network.
>
> What happens if somebody else tries to log on just after a reboot?
That too is fine; any other user, either after a reboot or after the
frozen XFCE session has been taken out by "kill -15"
Jonathan
Well, then, the user should try to log in on a regular X11 session
(usual graphical login), get the freeze, then log in immediately via
ssh or a virtual console and look at his hidden files under the home
directory:
ls -lart ~/
or /home/username if the admin is checking for the user. These options
will show the most recently changed hidden directories, which may or
may not provide a clue. Just for grins, check /var/log immediately
with the same options. You might see some new logs with some clues
from the stalled login.
One hidden directory to check in would be ".cache" . Notice the
contents of .cache/sessions and .cache/menus .
Another hidden directory to check would be ".config" . There might be others.
How to tell what to delete? Go by sense of smell. Stuff in caches
should not be too much of a concern, but stuff in configurations, you
might want to check the man pages and think about once or twice first.
Anyway, try emptying the obvious caches first.
That is, kill the stalled session, empty a cache, try logging in again.
--
Joel Rees
8:26 p.m.
On 05/02/2012 05:30 PM, Joel Rees wrote:
Well, then, the user should try to log in on a regular X11 session
(usual graphical login), get the freeze, then log in immediately via
ssh or a virtual console and look at his hidden files under the home
directory:
ls -lart ~/
or /home/username if the admin is checking for the user. These options
will show the most recently changed hidden directories, which may or
may not provide a clue.
/home/user/.xsession-errors would be another good place to check.
6:06 a.m.
Joe Zeff said:
/home/user/.xsession-errors would be another good place to check.
Here is it - should that tell me how to go forward? The signal 15 is
when I terminated the xfce session from a root console session.
Jonathan
+++
/usr/bin/startxfce4: X server already running on display :0
ssh-agent is already running; starting gpg-agent without ssh support
xfdesktop[1438]: starting up
xfce4-settings-helper: Another instance is already running. Leaving...
Initializing tracker-miner-fs...
Tracker-Message: Setting up monitor for changes to config
file:'/home/tabitha/.config/tracker/tracker-miner-fs.cfg'
Starting log:
File:'/home/tabitha/.local/share/tracker/tracker-miner-fs.log'
GNOME_KEYRING_CONTROL=/tmp/keyring-gzQ2UL
GNOME_KEYRING_CONTROL=/tmp/keyring-gzQ2UL
SELinux Troubleshooter: Applet requires SELinux be enabled to run.
Initializing tracker-store...
** Message: applet now removed from the notification area
Tracker-Message: Setting up monitor for changes to config
file:'/home/tabitha/.config/tracker/tracker-store.cfg'
abrt-applet: glib < 2.31 - init threading
Tracker-Message: Setting up monitor for changes to config
file:'/home/tabitha/.config/tracker/tracker-store.cfg'
Starting log:
File:'/home/tabitha/.local/share/tracker/tracker-store.log'
GNOME_KEYRING_CONTROL=/tmp/keyring-gzQ2UL
GPG_AGENT_INFO=/tmp/keyring-gzQ2UL/gpg:0:1
(tracker-store:1541): Tracker-CRITICAL **: D-Bus service
name:'org.freedesktop.Tracker1' is already taken, perhaps the daemon is already
running?
GNOME_KEYRING_CONTROL=/tmp/keyring-gzQ2UL
GPG_AGENT_INFO=/tmp/keyring-gzQ2UL/gpg:0:1
SSH_AUTH_SOCK=/tmp/keyring-gzQ2UL/ssh
(xfce4-settings-helper:1569): libxfce4ui-WARNING **: ICE I/O Error
(xfce4-settings-helper:1569): libxfce4ui-WARNING **: Disconnected from session manager.
/etc/xdg/xfce4/xinitrc: line 221: 1426 Terminated xfce4-session
running 'pkill -INT ^gpg-agent$; rm -f /home/tabitha/.cache/gpg-agent-info'
xfsettingsd: Fatal IO error 11 (Resource temporarily unavailable) on X server :0.
Thunar: Fatal IO error 11 (Resource temporarily unavailable) on X server :0.0.
xfce4-settings-helper: Fatal IO error 11 (Resource temporarily unavailable) on X server
:0.0.
(polkit-gnome-authentication-agent-1:1447): Gdk-WARNING **:
polkit-gnome-authentication-agent-1: Fatal IO error 11 (Resource temporarily unavailable)
on X server :0.0.
g_dbus_connection_real_closed: Remote peer vanished with error: Underlying GIOStream
returned 0 bytes on an async read (g-io-error-quark, 0). Exiting.
Received signal:15
(nm-applet:1457): Gdk-WARNING **: nm-applet: Fatal IO error 11 (Resource temporarily
unavailable) on X server :0.0.
abrt-applet: Fatal IO error 11 (Resource temporarily unavailable) on X server :0.0.
ABRT service is not running
applet.py: Fatal IO error 11 (Resource temporarily unavailable) on X server :0.0.
g_dbus_connection_real_closed: Remote peer vanished with error: Underlying GIOStream
returned 0 bytes on an async read (g-io-error-quark, 0). Exiting.
xfwm4: Fatal IO error 2 (No such file or directory) on X server :0.0.
(xfdesktop:1438): libxfce4ui-WARNING **: ICE I/O Error
(xfdesktop:1438): libxfce4ui-WARNING **: Disconnected from session manager.
migrate: Fatal IO error 0 (Success) on X server :0.0.
No protocol specified
8:31 p.m.
On Thu, May 3, 2012 at 8:06 PM, Jonathan Allen
<jonathan(a)barumtrading.co.uk> wrote:
Joe Zeff said:
> /home/user/.xsession-errors would be another good place to check.
Here is it - should that tell me how to go forward? The signal 15 is
when I terminated the xfce session from a root console session.
Jonathan
+++
/usr/bin/startxfce4: X server already running on display :0
Hmm.
ssh-agent is already running; starting gpg-agent without ssh support
xfdesktop[1438]: starting up
xfce4-settings-helper: Another instance is already running. Leaving...
Huh?
Initializing tracker-miner-fs...
Tracker-Message: Setting up monitor for changes to config
file:'/home/tabitha/.config/tracker/tracker-miner-fs.cfg'
Starting log:
File:'/home/tabitha/.local/share/tracker/tracker-miner-fs.log'
GNOME_KEYRING_CONTROL=/tmp/keyring-gzQ2UL
GNOME_KEYRING_CONTROL=/tmp/keyring-gzQ2UL
SELinux Troubleshooter: Applet requires SELinux be enabled to run.
Initializing tracker-store...
** Message: applet now removed from the notification area
Tracker-Message: Setting up monitor for changes to config
file:'/home/tabitha/.config/tracker/tracker-store.cfg'
abrt-applet: glib < 2.31 - init threading
Tracker-Message: Setting up monitor for changes to config
file:'/home/tabitha/.config/tracker/tracker-store.cfg'
Starting log:
File:'/home/tabitha/.local/share/tracker/tracker-store.log'
GNOME_KEYRING_CONTROL=/tmp/keyring-gzQ2UL
GPG_AGENT_INFO=/tmp/keyring-gzQ2UL/gpg:0:1
(tracker-store:1541): Tracker-CRITICAL **: D-Bus service
name:'org.freedesktop.Tracker1' is already taken, perhaps the daemon is already
running?
Yeah, something is left over, here, too.
GNOME_KEYRING_CONTROL=/tmp/keyring-gzQ2UL
GPG_AGENT_INFO=/tmp/keyring-gzQ2UL/gpg:0:1
SSH_AUTH_SOCK=/tmp/keyring-gzQ2UL/ssh
(xfce4-settings-helper:1569): libxfce4ui-WARNING **: ICE I/O Error
(xfce4-settings-helper:1569): libxfce4ui-WARNING **: Disconnected from session manager.
/etc/xdg/xfce4/xinitrc: line 221: 1426 Terminated xfce4-session
At this point, you're stuck, I'm thinking.
running 'pkill -INT ^gpg-agent$; rm -f
/home/tabitha/.cache/gpg-agent-info'
There's a cache that you want to look at. See what's there, and either
delete it or move it to someplace like /home/isolationward/.cache .
Re-boot and see what happens.
xfsettingsd: Fatal IO error 11 (Resource temporarily unavailable) on
X server :0.
Thunar: Fatal IO error 11 (Resource temporarily unavailable) on X server :0.0.
xfce4-settings-helper: Fatal IO error 11 (Resource temporarily unavailable) on X server
:0.0.
(polkit-gnome-authentication-agent-1:1447): Gdk-WARNING **:
polkit-gnome-authentication-agent-1: Fatal IO error 11 (Resource temporarily unavailable)
on X server :0.0.
g_dbus_connection_real_closed: Remote peer vanished with error: Underlying GIOStream
returned 0 bytes on an async read (g-io-error-quark, 0). Exiting.
Received signal:15
(nm-applet:1457): Gdk-WARNING **: nm-applet: Fatal IO error 11 (Resource temporarily
unavailable) on X server :0.0.
abrt-applet: Fatal IO error 11 (Resource temporarily unavailable) on X server :0.0.
ABRT service is not running
applet.py: Fatal IO error 11 (Resource temporarily unavailable) on X server :0.0.
g_dbus_connection_real_closed: Remote peer vanished with error: Underlying GIOStream
returned 0 bytes on an async read (g-io-error-quark, 0). Exiting.
xfwm4: Fatal IO error 2 (No such file or directory) on X server :0.0.
(xfdesktop:1438): libxfce4ui-WARNING **: ICE I/O Error
(xfdesktop:1438): libxfce4ui-WARNING **: Disconnected from session manager.
migrate: Fatal IO error 0 (Success) on X server :0.0.
No protocol specified
--
I'm wondering if this is related to a problem I'm having where the
input method processes leak over a logout and then when I log in again
(any user), I have two input method icons in the task bar panel.
--
Joel Rees
10:10 p.m.
I had the disappearing title bar problem again just now and thought of
another thing you might try, referencing the thread I post a link to
several weeks ago:
http://lists.fedoraproject.org/pipermail/users/2011-December/411290.html
We used to set the default run level to 3 to keep X11 from starting,
or use telinit 3 to shut X11 down. I'm not sure how systemctld does
that now, but it looks like either the stop or isolate command is what
you would use instead of telinit, ...
(crud. I'm looking at man systemd.unit and I see something I'd
suspected, but hadn't seen before:
The syntax is inspired by XDG Desktop Entry Specification[1]
.desktop files,
which are in turn inspired by Microsoft Windows .ini files.
Who turned this guy loose in our source tree? Man, I wish I had enough
money to go back and fork Fedora from before he get allowed in. Even
just enough money to pay my rent and feed the kids and set up some
servers in the other room or something until I got it going.
Maybe it is time to bite the bullet and start moving my stuff to
Debian, or check out how far Cent has gone down this road.)
Anyway, if you don't want to take the time to look up the unit names
for X11 services or whatever he's calling that, I think telinit 3 will
still work in F16.
With X11 down, log in to the account for the user that is having
problems, and use startx or the current equivalent, and see if you can
start that way. If you can, get a virtual terminal and run
xfwm4 --replace &
Running that in a virtual console, without X11 running, probably
doesn't do much good.
Or, if you have xde or gnome installed, try logging in on one of those
instead, and run the above command, as a swag.
It should reset a number of defaults for you, which should be less
brutal than what I outlined below.
On Fri, May 4, 2012 at 10:31 AM, Joel Rees <joel.rees(a)gmail.com> wrote:
On Thu, May 3, 2012 at 8:06 PM, Jonathan Allen
<jonathan(a)barumtrading.co.uk> wrote:
> Joe Zeff said:
>> /home/user/.xsession-errors would be another good place to check.
>
> Here is it - should that tell me how to go forward? The signal 15 is
> when I terminated the xfce session from a root console session.
>
> Jonathan
> +++
>
> /usr/bin/startxfce4: X server already running on display :0
Hmm.
> ssh-agent is already running; starting gpg-agent without ssh support
> xfdesktop[1438]: starting up
> xfce4-settings-helper: Another instance is already running. Leaving...
Huh?
> Initializing tracker-miner-fs...
> Tracker-Message: Setting up monitor for changes to config
file:'/home/tabitha/.config/tracker/tracker-miner-fs.cfg'
> Starting log:
> File:'/home/tabitha/.local/share/tracker/tracker-miner-fs.log'
> GNOME_KEYRING_CONTROL=/tmp/keyring-gzQ2UL
> GNOME_KEYRING_CONTROL=/tmp/keyring-gzQ2UL
> SELinux Troubleshooter: Applet requires SELinux be enabled to run.
> Initializing tracker-store...
> ** Message: applet now removed from the notification area
> Tracker-Message: Setting up monitor for changes to config
file:'/home/tabitha/.config/tracker/tracker-store.cfg'
> abrt-applet: glib < 2.31 - init threading
> Tracker-Message: Setting up monitor for changes to config
file:'/home/tabitha/.config/tracker/tracker-store.cfg'
> Starting log:
> File:'/home/tabitha/.local/share/tracker/tracker-store.log'
> GNOME_KEYRING_CONTROL=/tmp/keyring-gzQ2UL
> GPG_AGENT_INFO=/tmp/keyring-gzQ2UL/gpg:0:1
>
> (tracker-store:1541): Tracker-CRITICAL **: D-Bus service
name:'org.freedesktop.Tracker1' is already taken, perhaps the daemon is already
running?
Yeah, something is left over, here, too.
> GNOME_KEYRING_CONTROL=/tmp/keyring-gzQ2UL
> GPG_AGENT_INFO=/tmp/keyring-gzQ2UL/gpg:0:1
> SSH_AUTH_SOCK=/tmp/keyring-gzQ2UL/ssh
>
> (xfce4-settings-helper:1569): libxfce4ui-WARNING **: ICE I/O Error
>
> (xfce4-settings-helper:1569): libxfce4ui-WARNING **: Disconnected from session
manager.
> /etc/xdg/xfce4/xinitrc: line 221: 1426 Terminated xfce4-session
At this point, you're stuck, I'm thinking.
> running 'pkill -INT ^gpg-agent$; rm -f /home/tabitha/.cache/gpg-agent-info'
There's a cache that you want to look at. See what's there, and either
delete it or move it to someplace like /home/isolationward/.cache .
Re-boot and see what happens.
> xfsettingsd: Fatal IO error 11 (Resource temporarily unavailable) on X server :0.
> Thunar: Fatal IO error 11 (Resource temporarily unavailable) on X server :0.0.
> xfce4-settings-helper: Fatal IO error 11 (Resource temporarily unavailable) on X
server :0.0.
>
> (polkit-gnome-authentication-agent-1:1447): Gdk-WARNING **:
polkit-gnome-authentication-agent-1: Fatal IO error 11 (Resource temporarily unavailable)
on X server :0.0.
>
> g_dbus_connection_real_closed: Remote peer vanished with error: Underlying GIOStream
returned 0 bytes on an async read (g-io-error-quark, 0). Exiting.
>
> Received signal:15
>
> (nm-applet:1457): Gdk-WARNING **: nm-applet: Fatal IO error 11 (Resource temporarily
unavailable) on X server :0.0.
>
> abrt-applet: Fatal IO error 11 (Resource temporarily unavailable) on X server :0.0.
> ABRT service is not running
> applet.py: Fatal IO error 11 (Resource temporarily unavailable) on X server :0.0.
> g_dbus_connection_real_closed: Remote peer vanished with error: Underlying GIOStream
returned 0 bytes on an async read (g-io-error-quark, 0). Exiting.
> xfwm4: Fatal IO error 2 (No such file or directory) on X server :0.0.
>
> (xfdesktop:1438): libxfce4ui-WARNING **: ICE I/O Error
>
> (xfdesktop:1438): libxfce4ui-WARNING **: Disconnected from session manager.
> migrate: Fatal IO error 0 (Success) on X server :0.0.
> No protocol specified
> --
I'm wondering if this is related to a problem I'm having where the
input method processes leak over a logout and then when I log in again
(any user), I have two input method icons in the task bar panel.
--
Joel Rees
--
--
Joel Rees
4372
days inactive
4387
days old
10 comments
3 participants
participants (3)
-
Joe Zeff -
Joel Rees -
Jonathan Allen