Errors with 6.0.78 on Ubuntu 16.04

[danslak]

Hello,

I’ve been using NoMachine 5.3 for some time, but with this annoying double-login problem:

https://forums.nomachine.com/topic/cannot-connect-to-physical-desktop-issue-ubuntu-16-04-running-mate-desktop

So thought I’d try the latest 6.0.78-1 to see if this was resolved. Unfortunately, this does not work at all for me.

I’m running MATE desktop on Ubuntu 16.04.3 LTS, which uses lightdm. Am trying to connect to physical desktop only. Host details:

# uname -a
Linux <hostname> 4.4.0-112-generic #135-Ubuntu SMP Fri Jan 19 11:48:36 UTC 2018 x86_64 x86_64 x86_64 GNU/Linux

# lsb_release -a
LSB Version:    core-9.20160110ubuntu0.2-amd64:core-9.20160110ubuntu0.2-noarch:security-9.20160110ubuntu0.2-amd64:security-9.20160110ubuntu0.2-noarch
Distributor ID: Ubuntu
Description:    Ubuntu 16.04.3 LTS
Release:        16.04
Codename:       xenial

I think sometimes X causes issues, so to be clear whats running there:

# ps -ef | grep X
root      2394  2343  5 Jan24 tty7     1-04:41:14 /usr/lib/xorg/Xorg -core :0 -seat seat0 -auth /var/run/lightdm/root/:0 -nolisten tcp vt7 -novtswitch
root     22609  2343  0 16:36 tty8     00:00:11 /usr/lib/xorg/Xorg -core :1 -seat seat0 -auth /var/run/lightdm/root/:1 -nolisten tcp vt8 -novtswitch
dslaybac 22939     1  0 Jan30 ?        01:45:10 /usr/bin/Xvfb :25 -screen 0 1280x1024x8

 

I first did a complete uninstall of the 5.3 NoMachine , including completely removing the directory /usr/NX. Then installed the new deb file to ensure had a fully clean install. After starting nxserver, I cannot connect from a client (Windows 10, latest NoMachine 6.0.78), instead I get the error:

The session negotiation failed.
Error: Agent was not initialized properly.

In /usr/NX/var/log, nxserver.log in particular is not happy with constant repeating of these messages:

2018-02-14 19:53:07 590.428 510 NXSERVER Starting LS 6.0.78 and services.
2018-02-14 19:53:07 590.563 510 NXSERVER System information: Ubuntu 16.04.3 LTS, standalone.
2018-02-14 19:53:09 671.872 510 NXSERVER WARNING! Process ‘/usr/NX/bin/nxexec –node –user lightdm –priority realtime –mode 0 –pid 25’ with pid ‘628/628’ finished with exit code 11 after 1,669 seconds.
2018-02-14 19:53:09 672.379 510 NXSERVER WARNING! 8832568C5D8D521F65748526576486D4 nxnode died with exit code 11.
2018-02-14 19:53:09 747.330 510 NXSERVER WARNING! Session 8832568C5D8D521F65748526576486D4 crashed on display 1.
2018-02-14 19:53:11 335.848 510 NXSERVER WARNING! Process ‘/usr/NX/bin/nxexec –node –user lightdm –priority realtime –mode 0 –pid 25’ with pid ‘741/741’ finished with exit code 11 after 1,580 seconds.

This repeats 8 times then:

2018-02-14 19:53:25 355.782 510 NXSERVER WARNING! 65DAFA6ADF7A93E4D58795099B7EAA43 nxnode died with exit code 11.
2018-02-14 19:53:25 434.603 510 NXSERVER WARNING! Session 65DAFA6ADF7A93E4D58795099B7EAA43 crashed on display 1.
2018-02-14 19:53:25 434.797 510 NXSERVER WARNING! Maximum errors for local session [localhost:4000] reached.

Then, after 10 second pause, that whole sequence repeats again, and again, and again….

And if I try to connect from the client, it reports this:

2018-02-14 19:54:17 527.495 2852 NXSERVER User ‘dslaybac’ logged in from ‘SOMEIP’ using authentication method password.
2018-02-14 19:54:24 921.033 2945 NXNODE WARNING! NXRunUnLog failed with error 2, No such file or directory.
2018-02-14 19:54:28 989.388 2852 NXSERVER User ‘dslaybac’ from ‘SOMEIP’ logged out.
2018-02-14 19:54:28 997.861 2852 NXSERVER WARNING! Process ‘/usr/NX/bin/nxexec –user –gssremove’ with pid ‘3120/3120’ finished with exit code 1 after 0,004 seconds.
2018-02-14 19:54:28 998.156 2852 NXSERVER ERROR! Cannot remove ticket. Error is:
2018-02-14 19:54:44 145.693 510 NXSERVER ERROR! Node for session 1 associated with pid 22609 was not started.
2018-02-14 19:54:45 774.137 510 NXSERVER WARNING! Process ‘/usr/NX/bin/nxexec –node –user lightdm –priority realtime –mode 0 –pid 12’ with pid ‘3212/3212’ finished with exit code 11 after 1,621 seconds.
2018-02-14 19:54:45 774.372 510 NXSERVER WARNING! 4D9FF360B8D3C6544C809D5608D7914A nxnode died with exit code 11.
2018-02-14 19:54:45 899.939 510 NXSERVER WARNING! Session 4D9FF360B8D3C6544C809D5608D7914A crashed on display 1.
2018-02-14 19:54:45 983.565 510 NXSERVER Shutting down NoMachine server and services.
2018-02-14 19:54:45 984.700 595 NXNODE WARNING! NXRunUnLog failed with error 2, No such file or directory.
2018-02-14 19:54:46 007.884 595 NXNODE WARNING! PA_SendCommandCli: Unable to open cli socket. Aborting command [set-default-source nx_voice_out.monitor] execution.
2018-02-14 19:54:46 320.846 510 NXSERVER WARNING! Process ‘/usr/NX/bin/nxexec –node –user lightdm –priority realtime –mode 0 –pid 12’ with pid ‘3321/3321’ finished with exit code 1 after 0,415 seconds.
2018-02-14 19:54:46 321.041 510 NXSERVER WARNING! 7D7ECDA023AC1B5EA0391B27AFDCD165 nxnode died with exit code 1.
2018-02-14 19:54:46 321.141 510 NXSERVER WARNING! Display not defined, drop start node cleaner.
2018-02-14 19:54:46 321.509 510 NXSERVER WARNING! Cannot set session as failed: 7D7ECDA023AC1B5EA0391B27AFDCD165 does no exist.

nxerror.log has alot of this:

2635 2635 19:54:10 359.928 Socket: WARNING! Error is 99 ‘Cannot assign requested address’.
2724 2724 19:54:12 040.805 Socket: WARNING! Can’t bind the IPv6 socket.
2724 2724 19:54:12 040.854 Socket: WARNING! Error is 99 ‘Cannot assign requested address’.
Info: Handler started with pid 2852 on Wed Feb 14 19:54:16 2018.
Info: Handling connection from SOMEIP port 64163 on Wed Feb 14 19:54:16 2018.
nxexecRemoveKerberosTicket: WARNING! Can’t remove ticket.
nxexecRemoveKerberosTicket: Error is No such file or directory.
2852 2852 19:54:24 237.537 Socket: WARNING! Can’t bind the IPv6 socket.
2852 2852 19:54:24 237.609 Socket: WARNING! Error is 99 ‘Cannot assign requested address’.
2945 2945 19:54:24 660.305 Socket: WARNING! Can’t bind the IPv6 socket.
2945 2945 19:54:24 660.362 Socket: WARNING! Error is 99 ‘Cannot assign requested address’.

I can collect the full logs and email in if helpful, but since this seems to be spitting out errors even before trying to connect, thought maybe something more fundamental is going on, and might be obvious from the above.

thank you,

Dan

 

 

[Mth]

Hello.

It seems that you have two problems here.

First one is:
The session negotiation failed.
Error: Agent was not initialized properly.

This we have reproduced and it is connected with the Xvfb process:
dslaybac 22939     1  0 Jan30 ?        01:45:10 /usr/bin/Xvfb :25 -screen 0 1280x1024x8

NoMachine doesn’t support 8bit color, so any attempts to attach to such desktop will always cause
agent not to be initialized. Changing to 24 bits should help here.

Second problem is that the nxexec –node –user lightdm process crashes with 11. We would need some more data to help with this. Can you check if the operating system is generating crash file for these and provide us with backtrace of core dump? Also the process is running with parameters:

‘/usr/NX/bin/nxexec –node –user lightdm –priority realtime –mode 0 –pid 12’

Can you tell us what is this process designated with –pid parameter – 12 in this case. We would need command line and environment for this process. Also can you confirm the operating system is in the login window state?

If not, it could be a cause for a problem as we are detecting that login window session is running here.

If there is no core dump for this case, please provide us with full logs to forum[at]nomachine[dot]com. You can find instructions about debug and collecting logs here: https://www.nomachine.com/DT07M00098.

/Mth

[danslak]

Ok, thank you! This set me on the right path, looking at X processes – it appears to be working fully correctly now for the first time since I switched to Ubuntu a few years ago!

Xvfb was not the culprit, although it did not need to be running. It was just a remnant from something I’d run a few weeks ago. (it would cause a X display option :25 to appear in nxplayer when you try to connect, along with root and user options, but the :25 one never got you anywhere).

I attempted to track down the process you mention (any pids associated with the nxexec line), but by the time I was looking for that, the problem disappeared. And I would need one of those crashes (after having saved a process listing) to see what pid was being referred to. So not quite sure what was going on there.

But in any case, there were two problems: light-locker (screen lock utility, which had been installed by default, either with Ubuntu or MATE desktop), and an IBM BigFix agent that is required to run on this machine. The BigFix agent has a process XBESClientUI, which apparently confuses nxserver. This is solved if you simply stop besclient, restart nxserver, then restart besclient (or I suppose if I fix the order they start after boot, this would be solved permanently); nxserver needs to start before besclient. Without that fix, you cannot login (maybe this was the source of the crashes?).

As for light-locker, once I removed that, the double-login problem disappeared. With that problem, when the machine screen locks, light-locker appears uses some different X display to provide the unlock screen (which looks like the initial login screen), and apparently this is a problem for nomachine because you first need to connect to that x display to unlock, but your desktop is on a different x display. At least, thats my interpretation: in that situation, after you connect the client/nxplayer, you then need to select the root X display (there’s a root option, and a user option), which brings you to the unlock screen, where you put in your password, but then the screen goes blank. You then need to close nxplayer, restart it, and then you’re logged in. But now, with light-locker uninstalled, I get a different unlock screen (mate-screensaver maybe?), and upon entering password, I’m immediately back to my desktop, no need to close nxplayer / restart.  HUGE improvement!

So key points:

1. uninstall light-locker

2. make sure nxserver starts AFTER besclient (IBM BigFix client). (or perhaps anything else that leaves a process running with an X in the name?)

Hopefully that is it and I’m good for now…

[Britgirl]

@Danslak, just to confirm that the TR which we opened thanks to your other forum post is in the queue – a solution should be made available in 6.1. (http://nomachine.com/TR11O08235)

[Author]

Posts