Server freezes in Windows 10 Enterprise when user connects

[fred1]

Hi,

I’ve installed NoMachine 6.0.66_2 for Windows in Windows 10 Enterprise and Windows 8.1 Pro. When I connect to the Windows 10 Enterprise machine from the Windows 8.1 Pro machine, the connection is established but approx 2 seconds after, the Windows 10 Enterprise machine freezes, and I see the connection drops on the client side (Windows 8.1 Pro). I then need to do a hard reboot on the Windows 10 Enterprise machine.

Anyone else had this issue and if so, did you find a fix?

Thanks,

Fred

[fred1]

EDIT: If I connect from the Windows 10 Enterprise machine to the Windows 8.1 Pro machine, it works perfectly fine. So it’s an issue with NoMachine in Windows 10 Enterprise. On Windows 10 Enterprise, I look in the user’s NoMachine’s log folder and it is empty. I guess that NoMachine freezes before being able to log anything. It freezes very soon after the connection is established. So NoMachine does something that Windows 10 Enterprise does not like after establishing a connection. I do have the time to see the Windows 10 desktop in my Windows 8.1 NoMachine client and I also do have the time to see on my Windows 10 display NoMachine’s “New user is connected” notification. But Windows 10 Enterprise freezes right after that, the notification is displayed in the freeze. No blue screen or anything, just a total freeze.

I have the latest build of Windows 10 Enterprise and it is all updated.

Thought I could give that information…

[fra81]

Hi Fred,

this issue is not known. Could you try to uncheck the ‘Use acceleration for display processing’ checkbox in Server preferences -> Performance tab? You can open Server preferences from the NoMachine tray icon by clicking on ‘Show the service status’, on your Windows 10 machine, before starting the connection.

[fred1]

Hi fra81,

I tested what you proposed (I believe I already tested it but anyway) and it still has the same bug.

I do have some logs however, I was not looking in the right folder. It is short so I will just copy and paste:

From nxserver.log:

2018-01-20 13:27:39 573.995  4668 NXNODE   WARNING! NXClient –monitor pid handle closed.

2018-01-20 13:27:39 573.995  4668 NXNODE   ERROR! NXClient Monitor failed to start.

2018-01-20 13:27:39 683.372  4668 NXNODE   WARNING! Cannot write to FD#1048.

2018-01-20 13:27:39 683.372  4668 NXNODE   WARNING! Error is: 9, ‘Bad file descriptor’.

2018-01-20 13:29:46 466.476  4960 NXSERVER Shutting down NoMachine server and services.

2018-01-20 13:29:46 689.129  4960 NXSERVER WARNING! Process ‘C:\WINDOWS\\System32\\cmd.exe cmd /C “C:\WINDOWS\\System32\\tasklist.exe” /fi “USERNAME eq nx”‘ with pid ‘6880/1068′ finished with exit code 1 after 0,161 seconds.

2018-01-20 13:30:31 851.399  5728 NXSERVER Starting WS 6.0.66 and services.

2018-01-20 13:30:31 851.399  5728 NXSERVER System information: Windows 10, standalone.

2018-01-20 13:32:51 289.080  7584 NXSERVER ERROR! Process with pid 5728 is not running.

2018-01-20 13:32:51 289.080  7584 NXSERVER ERROR! Server process with pid 5728 finished without correct cleaning for session id ’78E1B03C6C6CF11862999078CD6CA475’, type ‘physicalDesktop’, owner ‘Fred’.

2018-01-20 13:32:51 320.330  7584 NXSERVER Starting WS 6.0.66 and services.

2018-01-20 13:32:51 320.330  7584 NXSERVER System information: Windows 10, standalone.

From nxservice.log:

2840 6972 13:29:46 468.272 System: WARNING! No object found to forward the signal 30, SIGPWR.

I believe that 13:29:46 is when the Windows 10 Ent box freezes and I need to reboot. 13:30:31 is when it has rebooted. It seems like the server cannot write to FD#1048. Maybe I have a permission issue? There’s also this NXClient failed to start… I obviously know nothing about NoMachine’s architecture so I don’t know if those errors explain the freeze I am getting when I try to connect…

I looked at other logs but they either had nothing logged for the correct timeframe or could not see anything that looked wrong.

Hope all this explains it.

Cheers,

Fred

[fred1]

EDIT: Full server-side debug logs (level 7) included. Connection happens at 14:06:27 and then box freezes. I believe there are a lot of blanks in the logs when the box is frozen.

Attachments:

1. log.zip

[fra81]

Hi fred1,

we’ve never seen anything similar. Indeed log files look zeroed when the freeze occurs. Maybe you could run a sanity check on your hard disk, just to be sure.

Please provide the following information:

1. Gather content of “C:\Windows\Minidump” directory.

2. Export Event Viewer logs:

– Open Control Panel -> System and Security -> Administrative Tools -> Event Viewer

– To export events, select the type of logs you want to export, click on Action menu and pick Save All Events As.

We’re interested in ‘Windows Logs/System’, ‘Windows Logs/Application’ and ‘Application and Services Logs/Hardware Events’ logs.

[fred1]

Hi,

Sent to the forum’s email.

Btw:

1. No minidumps at all. And I checked to make sure my Windows 10 was configured to generate some (it is always by default but I checked anyway).

2. Your events are sent by email. There were no hardware events at all. However, I could see an error with the creation of the nx user. Might be the culprit. Don’t know.

Since it is a very stable bug with NoMachine on my Windows 10 Ent box, always at the same moment, I believe it is not HDD related. I did a chkdsk, nothing.

Thanks,

Fred

[Cato]

Hello fred1,

Installation log file is missing from logs you previously sent us. File can be found in “C:\Users\<username>\AppData\Local\Temp”, where username denotes user account used during installation. File name format is “Setup Log <date>.txt”. In case of multiple Setup Log files, send us the most recent.

You can also check if your system correctly creates new user accounts. Create a new account and log on user’s desktop.

[Author]

Posts