I am trying to get ubuntu-desktop to work on an AWS instance using the in ‘NVIDIA Gaming PC - Ubuntu 18.04’ Amazon Machine Image. The AMI is installed on a g4dn.xlarge instance, which has an Nvidia T4. I followed the instructions at Running Ubuntu Desktop on an AWS EC2 instance | Ubuntu about installing Ubuntu desktop on an AWS system using TightVNC. I can establish a VNC connection to the AMI, but the desktop is only partly functional; there is no Gnome side panel, some applications do not start at all, and windows do not have minimize-maximize-close buttons, for example.
nvidia-smi works correctly, but nvidia-settings returns ‘ERROR: Unable to load info from any available system’. I have tried some suggestions I found in this forum, with no luck, and I have pretty much exhausted my ability to trace through all the pieces of a desktop GUI.
I have a debug log file but I can’t find the “upload” button I was expecting to see after I wrote this. I will keep looking…
The upload button is in the row of formatting options while creating a post (horizontal bar with upwards arrow).
The howto is about running a vncserver which is only a software emulated Xserver. For running nvidia-settings (or any hw accellerated gphics software) on the nvidia gpu, you will have to start an Xserver on it (the real Xserver is currently running on the vm gfx) and the connect to it, e.g. through x11vnc.
To run and Xserver on the nvidia, please create an xorg.conf. You could try running nvidia-x-config to have it created automatically.
-
When I try to run nvidia-x-config it says ‘command not found’ and suggests apt install nvidia-340 or apt install nvidia-utils-390. I am using a 470.74 driver. Will nvidia-utils-390 work, or should I look for an nvidia-utils-470.74?
-
It sounds like I should remove vncserver and instead use x11vnc; is that right?
Instead of searching for nvidia-x-config, just create /etc/X11/xorg.conf containing just
Section "Device"
Identifier "nvidia"
Driver "nvidia"
VendorName "NVIDIA Corporation"
BusID "PCI:30:0:0"
Option "AllowEmptyInitialConfiguration"
EndSection
and reboot. Afterwards, please create a new nvidia-bug-report.log
If you installed tigervnc to get vncserver, it should also contain x0vncserver which is the same as x11vnc, so you can use that.
Since the standard gdm is spawning a new Xserver on login, you should chnage to e.g. lightdm instead.
This is really helpful, thank you. I installed tightvncserver, so will remove that and install tigervnc for x0vncserver. This is going to take me a while because I am a cautious plodder and it’s getting late in the work day, but will update tomorrow with new log file.
The package name is tigervnc-scraping-server
I have installed tigervnc, changed from gdm to lightdm, and created that /etc/X11/xorg.conf, and I can’t get startx to work. I think this is the relevant part of the terminal output and log file:
$ startx – :1 -logverbose 6
X.Org X Server 1.19.6
Release Date: 2017-12-20
X Protocol Version 11, Revision 0
Build Operating System: Linux 4.15.0-140-generic x86_64 Ubuntu
Current Operating System: Linux ip-172-31-27-160 5.4.0-1058-aws #61~18.04.3-Ubuntu SMP Fri Oct 1 14:04:01 UTC 2021 x86_64
…
xorg-server 2:1.19.6-1ubuntu4.9 (For technical support please see Enterprise open source support | Ubuntu)
Current version of pixman: 0.34.0
Before reporting problems, check http://wiki.x.org
to make sure that you have the latest version.
Markers: (–) probed, (**) from config file, (==) default setting,
(++) from command line, (!!) notice, (II) informational,
(WW) warning, (EE) error, (NI) not implemented, (??) unknown.
(==) Log file: “/var/log/Xorg.1.log”, Time: Tue Oct 26 16:16:04 2021
(==) Using config file: “/etc/X11/xorg.conf”
(==) Using config directory: “/etc/X11/xorg.conf.d”
(==) Using system config directory “/usr/share/X11/xorg.conf.d”
(EE)
Fatal server error:
(EE) no screens found(EE)
(EE)
Please consult the The X.Org Foundation support
at http://wiki.x.org
for help.
(EE) Please also check the log file at “/var/log/Xorg.1.log” for additional information.
(EE)
(EE) Server terminated with error (1). Closing log file.
xinit: giving up
xinit: unable to connect to X server: Connection refused
xinit: server error
Couldn’t get a file descriptor referring to the console
ubuntu@ip-172-31-27-160:~$ tail -25 /var/log/Xorg.1.log
[ 8650.334] ABI class: X.Org Video Driver, version 23.0
[ 8650.334] (EE) open /dev/fb0: No such file or directory
[ 8650.334] (WW) Falling back to old probe method for vesa
[ 8650.334] (WW) Falling back to old probe method for modesetting
[ 8650.334] (WW) Falling back to old probe method for fbdev
[ 8650.334] (II) Loading sub module “fbdevhw”
[ 8650.334] (II) LoadModule: “fbdevhw”
[ 8650.335] (II) Loading /usr/lib/xorg/modules/libfbdevhw.so
[ 8650.335] (II) Module fbdevhw: vendor=“X.Org Foundation”
[ 8650.335] compiled for 1.19.6, module version = 0.0.2
[ 8650.335] ABI class: X.Org Video Driver, version 23.0
[ 8650.335] (EE) open /dev/fb0: No such file or directory
[ 8650.335] (WW) Falling back to old probe method for vesa
[ 8650.335] (EE) [drm] Failed to open DRM device for (null): -2
[ 8650.335] (EE) No devices detected.
[ 8650.335] (EE)
Fatal server error:
[ 8650.335] (EE) no screens found(EE)
[ 8650.335] (EE)
Please consult the The X.Org Foundation support
at http://wiki.x.org
for help.
[ 8650.335] (EE) Please also check the log file at “/var/log/Xorg.1.log” for additional information.
[ 8650.335] (EE)
[ 8650.388] (EE) Server terminated with error (1). Closing log file.
The full Xorg.1.log is attached.
I may be confused about how to start the VNC server (or something else). I am trying to run startx by hand because when I run x0vncserver it says “unable to open display :1”, so I assume something else needs to start up an X server.
Xorg.1.log (10.4 KB)
Sorry, I had a typo in the xorg.conf
BusID "PCI:30:0:0"
should read
BusID "PCI:0:30:0"
Well, I’ll be darned, startx works now and x0vncserver starts and says “listening on port 5900”, so it seems to be working on that end now. VNC Viewer times out waiting for a connection, which didn’t happen before, but I don’t think I can blame that on you. Maybe some sort of Amazon Cloud security rule.
I say you fixed my problem. Thank you very much.