Constant Lost Sessions

ThinLinc
1

Hello: I was using ThinLinc without issue for a few weeks just in home: server and client were both Ubuntu 22.04. The server had to be reflashed, so I moved to Debian 12. Since then, ThinLinc successfully connects, but then inevitably loses the connection. And then when I connect a monitor to the server, to restart it, my user appears to be logged in twice.

I found the bug associated with the gnome-screensaver, so removed that from, but no change. Not sure what to try next. I was surprised not to find any posts here about this issue.

Thanks for any guidance.

2

Hello @eumaios,

What do you mean more specifically with “loses the connection”? Could you provide the session log for such a scenario (/var/opt/thinlinc/sessions//last/xinit.log. And what information does the ThinLinc client provide in this situation?

my user appears to be logged in twice.

What does this mean, can you provide more information, please?

Kind regards,
Martin

3

Hi, Martin:

So shortly after establishing a connection, I lose all access to the server. I can see it, but have no control (screen’s frozen: clock stops). Then not long after this, I will get the attached message.

When I connect a monitor to the Debian server, it displays twice the user with which I logged in via ThinLinc–never seen that before (didn’t even know the same user could be logged in twice).

thinlinc

Below is the xinit.log you requested.

Thank you!

xauth: file /var/opt/thinlinc/sessions/privacy/10/Xauthority does not exist

Xvnc ThinLinc release 4.15.0 build 3358 - built Aug 21 2023 07:22:38
Copyright (C) 1999-2023 TigerVNC Team and many others (see README.rst)
See https://www.tigervnc.org for information on TigerVNC.
Underlying X server release 12013000

(II) Initializing extension VNC-EXTENSION

Tue Sep 12 06:15:52 2023
vncext: VNC extension running!
vncext: Listening for VNC connections on local interface(s), port 5910
vncext: created VNC server for screen 0
(II) Initializing extension Generic Event Extension
(II) Initializing extension SHAPE
(II) Initializing extension MIT-SHM
(II) Initializing extension XInputExtension
(II) Initializing extension SHAPE
(II) Initializing extension MIT-SHM
(II) Initializing extension XInputExtension
(II) Initializing extension XTEST
(II) Initializing extension BIG-REQUESTS
(II) Initializing extension SYNC
(II) Initializing extension XKEYBOARD
(II) Initializing extension XC-MISC
(II) Initializing extension XFIXES
(II) Initializing extension RENDER
(II) Initializing extension RANDR
(II) Initializing extension COMPOSITE
(II) Initializing extension DAMAGE
(II) Initializing extension MIT-SCREEN-SAVER
(II) Initializing extension DOUBLE-BUFFER
(II) Initializing extension RECORD
(II) Initializing extension DPMS
(II) Initializing extension Present
(II) Initializing extension DRI3
(II) Initializing extension X-Resource
(II) Initializing extension XVideo
(II) Initializing extension XVideo-MotionCompensation
(II) Initializing extension GLX
(II) IGLX: Loaded and initialized swrast
(II) GLX: Initialized DRISWRAST GL provider for screen 0
The XKEYBOARD keymap compiler (xkbcomp) reports:

Warning: Unsupported high keycode 372 for name ignored
X11 cannot support keycodes above 255.
This warning only shows for the first high keycode.
Errors from xkbcomp are not fatal to the X server
Connections: accepted: 127.0.0.1::42280
VNCSConnST: closing 127.0.0.1::42280: Clean disconnection
EncodeManager: Framebuffer updates: 0
EncodeManager: Total: 0 rects, 0 pixels
EncodeManager: 0 B (1:-nan ratio)
Connections: closed: 127.0.0.1::42280
ComparingUpdateTracker: 0 pixels in / 0 pixels out
ComparingUpdateTracker: (1:-nan ratio)
tl-xinit: Xserver ready for clients.

— System information —
ThinLinc 4.15.0
Linux eumaios 6.1.0-12-amd64 #1 SMP PREEMPT_DYNAMIC Debian 6.1.52-1 (2023-09-07) x86_64 GNU/Linux
Debian GNU/Linux 12 \n \l

— End system information —
Fontconfig warning: “/usr/share/fontconfig/conf.avail/05-reset-dirs-sample.conf”, line 6: unknown element “reset-dirs”
Fontconfig warning: “/usr/share/fontconfig/conf.avail/05-reset-dirs-sample.conf”, line 6: unknown element “reset-dirs”
Connections: accepted: 127.0.0.1::42296
SConnection: Client needs protocol version 3.8
SConnection: Client requests security type VncAuth(2)
VNCSConnST: Replacing existing connection
VNCSConnST: Server default pixel format depth 24 (32bpp) little-endian rgb888
Running /opt/thinlinc/etc/xstartup.d/01-tl-kinit.sh (Obtaining Kerberos ticket)
Running /opt/thinlinc/etc/xstartup.d/02-tl-dbus-launch.sh (Starting session D-Bus server)
Existing session D-Bus server found.
Setting initial D-Bus and systemd environment…
Running /opt/thinlinc/etc/xstartup.d/03-tl-x11-allow-localuser
localuser:privacy being added to access control list
Running /opt/thinlinc/etc/xstartup.d/11-tl-set-title (Setting client window title)
Running /opt/thinlinc/etc/xstartup.d/12-tl-default-keyboard (Configuring default keyboard layout)
Setting default keyboard layout to “us”
The XKEYBOARD keymap compiler (xkbcomp) reports:

Warning: Unsupported high keycode 372 for name ignored
X11 cannot support keycodes above 255.
This warning only shows for the first high keycode.
Errors from xkbcomp are not fatal to the X server
Running /opt/thinlinc/etc/xstartup.d/15-tl-shadow-notify (Enabling notifications when the session is shadowed)
Running /opt/thinlinc/etc/xstartup.d/20-tl-select-profile.sh (Choosing a profile)
Running /opt/thinlinc/etc/xstartup.d/30-tl-xdg-user-dirs (Translating standard folders)
Running /opt/thinlinc/etc/xstartup.d/40-tl-mount-localdrives (Mounting local drives)
Running /opt/thinlinc/etc/xstartup.d/41-tl-lp-redir-launch (Setting up printer port redirection)
Running /opt/thinlinc/etc/xstartup.d/43-tl-pulseaudio-launch.sh (Starting session PulseAudio server)
W: [pulseaudio] main.c: Couldn’t canonicalize binary path, cannot self execute.
Session PulseAudio server started. Redirecting applications…
Running /opt/thinlinc/etc/xstartup.d/50-tl-wait-smartcard (Waiting for smart card connection)
Stopping initial window manager…
Done.
Executing profile: gnome
Using XDG session: gnome
Updating D-Bus and systemd environment…
Executing XDG session command: /usr/bin/gnome-session

Tue Sep 12 06:16:00 2023
ComparingUpdateTracker: 8.42564 Mpixels in / 3.72683 Mpixels out
ComparingUpdateTracker: (1:2.26081 ratio)
The XKEYBOARD keymap compiler (xkbcomp) reports:

Warning: Unsupported high keycode 372 for name ignored
X11 cannot support keycodes above 255.
This warning only shows for the first high keycode.
Errors from xkbcomp are not fatal to the X server
The XKEYBOARD keymap compiler (xkbcomp) reports:
Warning: Unsupported high keycode 372 for name ignored
X11 cannot support keycodes above 255.
This warning only shows for the first high keycode.
Errors from xkbcomp are not fatal to the X server
The XKEYBOARD keymap compiler (xkbcomp) reports:
Warning: Unsupported high keycode 372 for name ignored
X11 cannot support keycodes above 255.
This warning only shows for the first high keycode.
Errors from xkbcomp are not fatal to the X server
The XKEYBOARD keymap compiler (xkbcomp) reports:
Warning: Unsupported high keycode 372 for name ignored
X11 cannot support keycodes above 255.
This warning only shows for the first high keycode.
Errors from xkbcomp are not fatal to the X server

Tue Sep 12 06:16:32 2023
ComparingUpdateTracker: 19.6044 Mpixels in / 4.31043 Mpixels out
ComparingUpdateTracker: (1:4.54814 ratio)

Tue Sep 12 06:16:35 2023
ComparingUpdateTracker: 6.21811 Mpixels in / 3.59578 Mpixels out
ComparingUpdateTracker: (1:1.72928 ratio)

4

Could you provide some screenshot of what this is?

No route to host

This happens because something changed on the network between your client and server, and the ThinLinc client can no longer reach the server. This is not an issue with ThinLinc itself, but on the network side of things. Did your client or server suddenly switch to another network for some reason?

Regards,
Martin

5

So it isn’t showing the two users anymore. No idea what to make of that.

Yes, I figured that I’m losing the connection, but I didn’t have this problem prior to the ThinLinc update and switching to a Debian vs. Ubuntu server. So the question is which change/variable is responsible.

My first thought was that it was a sleep/hibernation setting in Debian. But I disabled these, and no change. So if it is an issue with Debian, I’m at a loss as to what it could be. Thus my inquiry here.

Are there any changes with the update that might underlie this?

My next attempt is going to be with ssh configs. Maybe it’s dropping the connection, though, again, I can’t figure out why it would start now.

Any further suggestions greatly appreciated. As I said, ThinLinc worked like a dream in contrast to other remote apps I’ve used. I hope to get back to that :slight_smile:

6

Have you figured out if this is something that happens on a regular interval? I.e. every X minutes or is it completly at random?

Could I ask what desktop environment and Display Manager you’re using on the local Debian 12 console?

Regards,
Martin

7

Hey, sorry for the delay responding; it’s been a busy week. I was able to solve the problem by tweaking my ssh configs. It seems the ssh connection was basically timing out and closing itself.

Not sure why this change was necessary, though. Because I was connecting to a Debian server rather than Ubuntu; or because of the ThinLinc update (only two variables, and both changed together).

Anyhow, thanks! Delighted I got this working smoothly again!

8

No problem. Would you mind sharing what parameters you changed, what the old values were and what the new are?

Kind regards

9

I played around quite a but, so this is as best I can recall. In sshd_config I believe I uncommented and gave the following values:
MaxAuthTries 3
MaxSessions 2
ClientAliveCountMax 4

As a followup question, could you tell the port numbers for the following which are set on install: tlwebaccess, tlwebadm, tlmaster, tlagent

Thanks!

10

You can read about which ports ThinLinc uses in our Administrators Guide

https://www.cendio.com/resources/docs/tag/tcp-ports.html?highlight=ports

Webaccess uses 300, webadm 1010, vsmserver 9000 and vsmagent 904

I don’t see how tweaking those settings in sshd would help a No route to host error. That would typically indicate a more low-level error on your network, and would not be limited to a specific service port on the remote machine.

Regards,
Martin