Dear list,
I run a small cluster for my lab, including about 12 Linux boxes running Ubuntu 14.04 (1x with 16.04). We have been using x2go and ssh as the primary tool to access these computing resources.
Recently (started a month ago), I noticed there were package updates of x2go server (now 4.1.0.0-0-1517), client (now 4.1.1.1-0-1719) and the nx11 libraries, on the other hand, I started having difficulty log on to some of my servers, and this list starts to grow. I am not sure if this was the result of the recent package upgrade, or the reason motivated the package updates, but I started seeing the below two types of messages:
For some of the machines, I saw the below message initially:
*The remote proxy closed the connection while negotiating*
**
*the session. This may be due to the wrong authentication*
**
*credentials passed to the server.*
for some other servers, I am getting the below message:
*Connection failed.**Failed to start X2Go Agent session with ID account-xxxxxx_stDXFCE_dp24. X2Go Agent terminated unexpectedly. Aborting session startup. *
The 1st message appeared first, after I tried to restart x2goserver or rebooting
the machine, or apply the updates, deleting the ~/.x2go folder, or a reboot, those
machines typically display the 2nd message; none of those were fixed using the
above procedures.
right now, I have about 2/3 of the my servers are not accessible via x2go.
I would like to know if there is a workaround. thanks
Qianqian
- On 03/26/2018 03:45 PM, Qianqian Fang wrote:
I run a small cluster for my lab, including about 12 Linux boxes running Ubuntu 14.04 (1x with 16.04). We have been using x2go and ssh as the primary tool to access these computing resources.
Recently (started a month ago), I noticed there were package updates of x2go server (now 4.1.0.0-0-1517), client (now 4.1.1.1-0-1719) and the nx11 libraries, on the other hand, I started having difficulty log on to some of my servers, and this list starts to grow. I am not sure if this was the result of the recent package upgrade, or the reason motivated the package updates, but I started seeing the below two types of messages:
Please make sure that nx-libs 3.5.99.16 is installed on the server side.
For some of the machines, I saw the below message initially:
*The remote proxy closed the connection while negotiating* ** *the session. This may be due to the wrong authentication* ** *credentials passed to the server.*
This sounds like a typical incarnation of the DPMS bug, fixed with nx-libs 3.5.99.15.
for some other servers, I am getting the below message:
*Connection failed.**Failed to start X2Go Agent session with ID account-xxxxxx_stDXFCE_dp24. X2Go Agent terminated unexpectedly. Aborting session startup. *
The 1st message appeared first, after I tried to restart x2goserver or rebooting
the machine, or apply the updates, deleting the ~/.x2go folder, or a reboot, those
machines typically display the 2nd message; none of those were fixed using the
above procedures.
However, this doesn't sound like it. Normally the DPMS bug only manifests after a pretty long time (say, an hour or 1.5h, depending on DPMS settings), but not right away after starting the session. I haven't yet seen such an error message yet. It's possible that it's caused by the session repositioning on non-first displays/nxagent crashing bug as fixed with nx-libs 3.5.99.16, but normally that incarnates differently (i.e., nxagent is being started successfully, but then dies quickly while starting the session without a client-side error message).
Mihai
On 03/26/2018 03:53 PM, Mihai Moldovan wrote:
Please make sure that nx-libs 3.5.99.16 is installed on the server side. This sounds like a typical incarnation of the DPMS bug, fixed with nx-libs 3.5.99.15.
thanks Mihai
I confirm that the server giving the first error, i.e. "The remote proxy closed the connection while negotiating..." has indeed installed an older version of libnx-x11 (3.5.0.33~710).
after letting apt-get to upgrade the nx libraries and x2go packages, I saw the libnx-x11 was replaced by a new package named libnx-x11-6, and the version is now 3.5.99.16. However, now I start seeing the 2nd error message on those machines, i.e. "connection failed".
However, this doesn't sound like it. Normally the DPMS bug only manifests after a pretty long time (say, an hour or 1.5h, depending on DPMS settings), but not right away after starting the session. I haven't yet seen such an error message yet. It's possible that it's caused by the session repositioning on non-first displays/nxagent crashing bug as fixed with nx-libs 3.5.99.16, but normally that incarnates differently (i.e., nxagent is being started successfully, but then dies quickly while starting the session without a client-side error message).
here is a dump of the session.log file under the failed session:
|NXAGENT - Version 3.5.99.16|| || ||Copyright (c) 2001, 2011 NoMachine (http://www.nomachine.com)|| ||....|| ||See https://github.com/ArcticaProject/nx-libs for more information.|| || ||Info: Agent running with pid '17844'.|| ||Session: Starting session at 'Mon Mar 26 16:56:11 2018'.|| ||Error: Aborting session with 'Unable to open display 'nx/nx,options=/tmp/.x2go-fangq/C-fangq-50-1522097770_stDXFCE_dp24/options:50''.|| ||Session: Aborting session at 'Mon Mar 26 16:56:11 2018'.|| ||Session: Session aborted at 'Mon Mar 26 16:56:11 2018'.|| |
strangely, the x2go window reports something differently, it says (see attached screenshot)
|"PANIC!" Failure negotiating the session in stage 7|| ||...|| ||Wrong version or invalid session authentication cookie.|| ||....|| |
are these two error messages caused by a single issue or multiple?
I've already restarted x2gosever, but no dice. any other files that I need to reset/clear?
thanks
Qianqian
Mihai
- On 03/26/2018 11:11 PM, Qianqian Fang wrote:
I confirm that the server giving the first error, i.e. "The remote proxy closed the connection while negotiating..." has indeed installed an older version of libnx-x11 (3.5.0.33~710).
Hmm, mixing X2Go Server 4.1.0.0 and nx-libs 3.5.0.x is not really supported.
after letting apt-get to upgrade the nx libraries and x2go packages, I saw the libnx-x11 was replaced by a new package named libnx-x11-6, and the version is now 3.5.99.16. However, now I start seeing the 2nd error message on those machines, i.e. "connection failed".
So, good that you upgraded.
here is a dump of the session.log file under the failed session:
|NXAGENT - Version 3.5.99.16|| || ||Copyright (c) 2001, 2011 NoMachine (http://www.nomachine.com)|| ||....|| ||See https://github.com/ArcticaProject/nx-libs for more information.|| || ||Info: Agent running with pid '17844'.|| ||Session: Starting session at 'Mon Mar 26 16:56:11 2018'.|| ||Error: Aborting session with 'Unable to open display 'nx/nx,options=/tmp/.x2go-fangq/C-fangq-50-1522097770_stDXFCE_dp24/options:50''.|| ||Session: Aborting session at 'Mon Mar 26 16:56:11 2018'.|| ||Session: Session aborted at 'Mon Mar 26 16:56:11 2018'.|| |
Hmm, strange. Being unable to open the internal display normally means there is a libNX_X11 problem.
dpkg -l | grep -Ei '(x2go|nx)' ls -ldh /usr/lib/*86*/nx/X11
strangely, the x2go window reports something differently, it says (see attached screenshot)
|"PANIC!" Failure negotiating the session in stage 7|| ||...|| ||Wrong version or invalid session authentication cookie.|| ||....|| |
are these two error messages caused by a single issue or multiple?
They are different error messages, because the first one is from nxproxy in client mode, while the second one is from nxproxy in server mode. This error message is a pretty generic one - the server part (which ironically runs on the client side for reasons) expected to read authentication data from the nxproxy client, but it never got it. Instead, the connection was just terminates unexpectedly.
I've already restarted x2gosever, but no dice. any other files that I need to reset/clear?
No. Additionally, you do not need to restart the x2goserver service at all (unless you're upgrading the x2goserver package or modifying /etc/x2go/x2goserver.conf) - the x2goserver service is just a cleanup utility that syncs content in the database with what is observed on the file system. It has nothing to do (well, mostly...) with starting or running sessions (though I've seen strange things happening leading to session suspension, but that's a different story and mainly caused by nxagent trying to display a popup or such).
Mihai
Qiangian,
Try removing the sessions on the server side and then see what happens on the client. I had a similar issue and ended up doing the package upgrades you've already done on the server end and then updating the client as well.
sudo rm -f /var/lib/x2go/x2go_sessions sudo x2godbadmin --createdb
Thanks, Steve
On 03/26/2018 04:11 PM, Qianqian Fang wrote:
On 03/26/2018 03:53 PM, Mihai Moldovan wrote:
Please make sure that nx-libs 3.5.99.16 is installed on the server side. This sounds like a typical incarnation of the DPMS bug, fixed with nx-libs 3.5.99.15.
thanks Mihai
I confirm that the server giving the first error, i.e. "The remote proxy closed the connection while negotiating..." has indeed installed an older version of libnx-x11 (3.5.0.33~710).
after letting apt-get to upgrade the nx libraries and x2go packages, I saw the libnx-x11 was replaced by a new package named libnx-x11-6, and the version is now 3.5.99.16. However, now I start seeing the 2nd error message on those machines, i.e. "connection failed".
However, this doesn't sound like it. Normally the DPMS bug only manifests after a pretty long time (say, an hour or 1.5h, depending on DPMS settings), but not right away after starting the session. I haven't yet seen such an error message yet. It's possible that it's caused by the session repositioning on non-first displays/nxagent crashing bug as fixed with nx-libs 3.5.99.16, but normally that incarnates differently (i.e., nxagent is being started successfully, but then dies quickly while starting the session without a client-side error message).
here is a dump of the session.log file under the failed session:
|NXAGENT - Version 3.5.99.16|| || ||Copyright (c) 2001, 2011 NoMachine (http://www.nomachine.com)|| ||....|| ||See https://github.com/ArcticaProject/nx-libs for more information.|| || ||Info: Agent running with pid '17844'.|| ||Session: Starting session at 'Mon Mar 26 16:56:11 2018'.|| ||Error: Aborting session with 'Unable to open display 'nx/nx,options=/tmp/.x2go-fangq/C-fangq-50-1522097770_stDXFCE_dp24/options:50''.|| ||Session: Aborting session at 'Mon Mar 26 16:56:11 2018'.|| ||Session: Session aborted at 'Mon Mar 26 16:56:11 2018'.|| |
strangely, the x2go window reports something differently, it says (see attached screenshot)
|"PANIC!" Failure negotiating the session in stage 7|| ||...|| ||Wrong version or invalid session authentication cookie.|| ||....|| |
are these two error messages caused by a single issue or multiple?
I've already restarted x2gosever, but no dice. any other files that I need to reset/clear?
thanks
Qianqian
Mihai
x2go-user mailing list x2go-user@lists.x2go.org https://lists.x2go.org/listinfo/x2go-user
--
Steven G. Spencer, Network Administrator KSC Corporate - The Kelly Supply Family of Companies Office 308-382-8764 Ext. 1131 Mobile 402-765-8010
-
Mihai Moldovan -
Qianqian Fang -
Steven Spencer