I am trying to figure out how I can VNC to my Ubuntu 19.04 desktop which doesn't have a monitor or keyboard/mouse. I installed Ubuntu 19.04 on an external USB hard drive using my MacBook Air. I then booted Ubuntu on the MacBook Air and installed SSH. Then the above hard drive which has Ubuntu installed was connected to my headless desktop and Ubuntu boots up perfectly and I was able to connect to my Ubuntu desktop via SSH on my MacBook Air. I installed tightvncserver on Ubuntu via SSH and when I connect to Ubuntu via VNC using MacBook Air, I get a grey screen on VNC. Read many articles and tutorials using xserver-xorg-video-dummy and also using dummy plugs for DVI and none of these works. How can I have Ubuntu boot up normally as if it was connected to a monitor and connect to it using VNC? Seems like Ubuntu 19.04 is not booting up the X server window without a monitor connected. Please don't provide solutions that requires using thrid-party desktop environments when Ubuntu already has a perfectly working desktop environment which only functions when monitor is connected. I booted Ubuntu 19.04 on my MacBook Air and logged in with the normal GUI interface. I started tightvncserver and when I login to VNC I also get a grey screen.

This is a pretty straightforward question. I am writing a Qt program that looks to render graphics in the terminal. To do that, I utilize Sixel Graphics which has a large overhead. I would rather render directly to the frame buffer, but can that be done while X and Wayland are running? Or do I have to strictly be in a headless environment, meaning no active desktop environment, to be able to access this? Does X or Wayland perhaps have functions allowing me to use it to render graphics through it? Thanks.

I am using Dreamhost VPS which doesn't provide sudo capability . Dreamhost runs on Ubuntu. My final goal is to be able to use Ruby to run headless web browser. I picked Watir gem and it requires chromedriver. I installed chromedriver into my home directory ~/chrome/, and added it to my $PATH. I get an error /home/user/.rvm/gems/ruby-2.3.0/gems/selenium-webdriver-3.141.0/lib/selenium/webdriver/remote/response.rb:69:in `assert_ok': unknown error: Chrome failed to start: exited abnormally (Selenium::WebDriver::Error::UnknownError) It seems I need to install Xvfb and all guides suggest sudo apt-get install, which I can't do. Even if I install Xvfb as standalone, it lacks a lot of extra stuff from X11: [myServer]$ Xvfb :1 -screen 0 1024x768x24 (WW) Could not open RGB file "/usr/X11R6/lib/X11/rgb.txt"; will use built-in copy. _XSERVTransSocketOpenCOTSServer: Unable to open socket for inet6 _XSERVTransOpen: transport open failed for inet6/ps554048:1 _XSERVTransMakeAllCOTSServerListeners: failed to open listener for inet6 error opening security policy file /usr/X11R6/lib/X11/xserver/SecurityPolicy Could not init font path element /usr/X11R6/lib/X11/fonts/misc/, removing from list! Could not init font path element /usr/X11R6/lib/X11/fonts/TTF/, removing from list! Could not init font path element /usr/X11R6/lib/X11/fonts/Speedo/, removing from list! Could not init font path element /usr/X11R6/lib/X11/fonts/Type1/, removing from list! Could not init font path element /usr/X11R6/lib/X11/fonts/CID/, removing from list! Could not init font path element /usr/X11R6/lib/X11/fonts/75dpi/, removing from list! Could not init font path element /usr/X11R6/lib/X11/fonts/100dpi/, removing from list! Fatal server error: could not open default font 'fixed' **Temporary solution** is to get instance on Amazon Web Services which allows 100% sudo and install all the necessary packages

I have a Windows pc that I use to connect to my Ubuntu PC over ssh that I use as a server (18.04) that is connected only to the internet and power (No monitor). When I run glxgears -info I only get 8-10 fps while having a graphics card. When I run it as xvfb-run glxgears -info I get 2000~ fps (Which I assume because it uses my GPU to render). But I can't get xvfb-run to display in my Xming X server. Running glxgears -info displays in Xming but as I said, low performance, etc. How Can I make xvfb to display in Xming on my Windows machine? Or is there an easier/better way to do this? Thanks!

I've a minipc Qotom Q1900G2-P with AlmaLinux 9.5.\ Qotom with American Megatrends bios, build 06.01.2015.\ Bios reset to factory, default setup. Without video cable, HDMI or VGA, it doesn't start...from dmesg or syslog or message NOTHING!\ It seems that even get to the grub menu but obviously not having video connected I do not see until the boot comes. If I connect the video later, with the minipc on, no video signal.\ Same with Debian 12.10. **OK with Windows!! :(( With Windows it start also without video cable!** I've trie to insert in /etc/default/grub some options as:\ GRUB_CMDLINE_LINUX_DEFAULT="quiet nomodeset console=tty1"\ or\ GRUB_CMDLINE_LINUX_DEFAULT="nomodeset text" But nothing! :(( some BIOS screen here https://imgur.com/a/TyDEDq5 Any ideas for solve? Thanks in advance

I submit defeat. I have been trying to configure my remote Linux box to have two dummy monitors so that I can use multiple local monitors to VNC into it. I'm surprised that no one else has needed help with this to find something on the web. I've also tried creating a monitor on the Linux box that is double wide. Then use x11vnc to -clip an area for each display. But I'm having issues creating a monitor that large with the dummy driver. I do have a graphics adapter installed that has two DisplayPorts but am not planning to use it. When I was using the real adapter, I was getting sluggish behavior. When I tried the dummy, it was very responsive. So I'm hoping to just create another dummy. I'm using KDE DM. I have seen many examples of using VIRTUAL1 but I can't get that working with the dummy driver. I tried adding Option "VirtualHeads" "2" into the config but the dummy driver doesn't recognize it. I’ve seen suggestions of using Xvfb but it has been deprecated by the dummy driver since 2016. Here are some details.

$ uname -a
Linux bgrupczy-linux 5.8.0-53-generic #60~20.04.1-Ubuntu SMP Thu May 6 09:52:46 UTC 2021 x86_64 x86_64 x86_64 GNU/Linux

/usr/share/X11/xorg.conf.d/xorg.conf is empty The following was gleaned from searching the internet. Once I got it working I stopped tweaking it so it may have flaws but that don't seem to affect me. /usr/share/X11/xorg.conf.d/dummy-1920x1080.conf has the following which gets me my single 1920x1080.

Section "Monitor"
  Identifier "Monitor0"
  HorizSync 28.0-80.0
  VertRefresh 48.0-75.0
  Modeline "1920x1080_60.00" 173.00 1920 2048 2248 2576 1080 1083 1088 1120 -hsync +vsync
EndSection

Section "Device"
  Identifier "Card0"
  Driver "dummy"
  VideoRam 256000
EndSection

Section "Screen"
  Identifier "Screen0"
  Device "Card0"
  Monitor "Monitor0"
  DefaultDepth 24
  SubSection "Display"
    Depth 24
    Modes "1920x1080_60.00"
  EndSubSection
EndSection

$ xrandr
xrandr: Failed to get size of gamma for output default
Screen 0: minimum 320 x 240, current 1920 x 1080, maximum 1920 x 1080
default connected 1920x1080+0+0 0mm x 0mm
   1920x1080     60.00* 
   1680x1050     70.00    60.00  
   1400x1050     70.00    60.00  
   1600x900      60.00  
   1280x1024     75.00    60.00  
   1440x900      60.00  
   1400x900      60.00  
   1280x960      60.00  
   1368x768      60.00  
   1360x768      60.00  
   1280x800      60.00  
   1152x864      75.00    70.00    60.00  
   1280x720      60.00  
   1024x768      75.00    70.00    60.00  
   1024x576      60.00  
   960x600       60.00  
   832x624       75.00  
   960x540       60.00  
   800x600       75.00    72.00    60.00    56.00  
   840x525       70.00    60.00  
   864x486       60.00  
   700x525       70.00    60.00  
   800x450       60.00  
   640x512       75.00    60.00  
   720x450       60.00  
   700x450       60.00  
   640x480       75.00    73.00    60.00  
   684x384       60.00  
   680x384       60.00  
   640x400       60.00  
   576x432       75.00    70.00    60.00  
   640x360       60.00  
   512x384       75.00    70.00    60.00  
   512x288       60.00  
   416x312       75.00  
   480x270       60.00  
   400x300       75.00    72.00    60.00    56.00  
   432x243       60.00  
   320x240       75.00    73.00    60.00

$ lspci | grep VGA
00:02.0 VGA compatible controller: Intel Corporation Device 9bc5 (rev 05)

I can see from xrandr that my monitor is called "default". So I tried this:

$ cvt 3840 1080
# 3840x1080 59.96 Hz (CVT) hsync: 67.16 kHz; pclk: 346.00 MHz
Modeline "3840x1080_60.00"  346.00  3840 4088 4496 5152  1080 1083 1093 1120 -hsync +vsync

$ xrandr --newmode "3840x1080_60.00"  346.00  3840 4088 4496 5152  1080 1083 1093 1120 -hsync +vsync
$ xrandr --addmode default "3840x1080_60.00"
$ xrandr --output default --mode "3840x1080_60.00"

The last line gives me:

xrandr: Configure crtc 0 failed

I tried to increase VideoRam to 512000 (double) to make sure I had room. Not sure what to do there. I have 32G ram. And ALL xrandr commands give me:

: Failed to get size of gamma for output default

and I figure that's because it's a dummy monitor that has no gamma. I'm currently Working within the VNC session. Do I need to shut down x11vnc to get xrandx to complete? I'm at my wits end. I'm starting x11vnc like this:

x11vnc -loop -forever -shared -repeat -noxdamage -xrandr -display :0 -clip 1920x1080+0+0

The linux box is within my local net so I'm not concerned with authentication/passwords. **Edit 2021-05-27:** More lurking and I found some options. I was able to get a double wide screen and then create two x11vnc instances. But this isn't optimal. The Linux box still sees this as a single screen. Now is there a way to take that screen and tell the Linux box to split it? For example, if I maximize a window in KDE it will not span both local screens? Xorg conf file:

Section "Monitor"
  Identifier "Monitor0"
EndSection

Section "Device"
  Identifier "Card0"
  Driver "dummy"
  VideoRam 512000
EndSection

Section "Screen"
  Identifier "Screen0"
  Device "Card0"
  Monitor "Monitor0"
  DefaultDepth 24
  SubSection "Display"
    Depth 24
    Virtual 3840 1080
  EndSubSection
EndSection

x11vnc -loop -forever -shared -repeat -noxdamage -xrandr -display :0 -rfbport 5900 -clip 1920x1080+0+0
x11vnc -loop -forever -shared -repeat -noxdamage -xrandr -display :0 -rfbport 5901 -clip 1920x1080+1920+0

I can then connect to VNC displays :0 and :1 and arrange them on local monitors and resize the windows to fit those monitors. When a dialog window appears, many times it's in the middle of the Linux "big screen" which for me spans both monitors... **Edit 2021-11-12:** One solution: https://superuser.com/a/1188573/514658 The real issue is not being able to get my VNC viewer to go full screen and only span two of my three monitors. The only option available in the viewer is to use ALL monitors which then blocks my use of the Windows side. DisplayFusion at above link did the trick. Now I can seamlessly drag between my two monitors and not get stuck on the edge of one. **Edit 2021-11-16:** "SOLVED": Instead of using DisplayFusion, which is a big hammer, more searching led me to Windows PowerShell. No need for a third party utility. I only need to move/resize a specific window to fit on two of my three monitors full screen. Here it is:

Add-Type @"
  using System;
  using System.Runtime.InteropServices;

  public class Win32 {
    [DllImport("user32.dll")]
    [return: MarshalAs(UnmanagedType.Bool)]
    public static extern bool GetWindowRect(IntPtr hWnd, out RECT lpRect);

    [DllImport("user32.dll")]
    [return: MarshalAs(UnmanagedType.Bool)]
    public static extern bool GetClientRect(IntPtr hWnd, out RECT lpRect);

    [DllImport("user32.dll")]
    [return: MarshalAs(UnmanagedType.Bool)]
    public static extern bool MoveWindow(IntPtr hWnd, int X, int Y, int nWidth, int nHeight, bool bRepaint);
  }

  public struct RECT
  {
    public int Left;        // x position of upper-left corner
    public int Top;         // y position of upper-left corner
    public int Right;       // x position of lower-right corner
    public int Bottom;      // y position of lower-right corner
  }

"@

$h = (Get-Process vnc-E4_6_3-x86_win32_viewer).MainWindowHandle

My issue is more nuanced than the title suggests. I have a pc I am using as a server. I first installed a headless ubuntu because that's all I needed to run my services. I then installed a standard ubuntu alongside the headless version to help me speed up the development of my services. My problem is this: If I set the boot to the headless version, as it boots, I don't have the option to boot into the headed version. If I set the boot to the headed version, then I have the option, but the default boot is the headed version. I want to know if I can A). add a boot manager for the headless version that times out and boots to the headless version by default, or B). edit the sequence of the boot manager for the headed version.

Is there a way I can have a process on a headless server which runs at startup, and still be able to connect remotely at a later time to view the console input? I need to be able to view console output via ssh or some other means. Simply redirecting output to a file won't work because these processes never exit and constantly output lines. If I just redirect to a file it'll fill up the disk... (I only really care about the last 20 lines or so.) I don't think "screen" is an option since I need run at boot time without interaction.

I have a Spotify player in JavaScript that I need to run headless in Raspian Bookworm on a RPi5. I am controlling the player using the spotipy module in python. I have both nginx and gunicorn running as system services, and everything is working when started in the GUI using VNC. However, I need this system to be headless and have it all start up on boot. I can start a Chromium browser in headless, but no sound is produced when starting a song in the Spotify player. Here is my service for the browser:

[Unit]
Description=Headfull Browser Service
After=gunicorn.service nginx.service network.target
Requires=gunicorn.service nginx.service

[Service]
ExecStartPre=/bin/sleep 10
ExecStart=/usr/bin/chromium-browser --headless --disable-gpu  --autoplay-policy=no-user-gesture-required --mute-audio=false --alsa-output-device=default --use-fake-ui-for-media-stream --remote-debugging-port=9222 http://localhost
Restart=always
User=myuser

[Install]
WantedBy=multi-user.target

I tried doing stuff with PulseAudio, but would rather not go that way. Is it possible to make this work in Chromium? Is there another easier way to run my JavaScript in headless?

asking for a friend: He has a UEFI laptop which he wants to install linux onto, the problem is that the integrated screen is broken, and as such he can't enter the bios, and change the boot priority to the linux USB installer. i had in mind of buying a RS-232 compatible adapter and plug it in a USB port, and try to control the bios from there, but i dont think the UEFI would display it's output in here, What could my friend do to access the UEFI interface without video output?

I'm using X11 forwarding to use GUI apps from my headless remote server. Everything was fine until I needed to use a program that needs system tray. Program needs a system tray to work correctly otherwise GUI doesn't show up. How can I run a GUI program that needs a system tray through X11 forwarding?

I'm working on a CentOS 6.3 x86_64 minimal OS (installed using the netinstall ISO). This system is not expected to run anything beyond console apps/services and Tomcat7. When I attempt to install OpenJDK 1.7, it lists some x11 packages as dependencies. For my purpose, not even awt is needed. And therefore, if possible, I would like to avoid installing all the fonts, x11, etc. Does there exist a Java/JRE version that comes without the UI pieces?

I would like to access a website using Firefox in Ubuntu from the terminal. I must do this *without opening a Firefox window* because my computer has no screen. How can I do this? Note: I am running an external program to capture Firefox traffic.

I have a fitlet2 (https://fit-iot.com/web/products/fitlet2/) which has a serial port. I can connect to the serial port of the fitlet2 with screen from my computer (screen /dev/ttyUSB0 115200) and access the BIOS, after I have enabled the "Serial Port Console Redirection" in the BIOS. I am also able to access the GRUB Bootloader over serial. Access over serial works in various GRUB bootloaders (debian/ubuntu installer, debian installed). But whenever the system is loaded, I can no more access the serial port which I was using. I tried to activate console redirection in GRUB for the system by editing the boot parameters according to this post https://www.cyberciti.biz/faq/howto-setup-serial-console-on-debian-linux/ . Soon I realized that my ttyS0 is somehow not working correctly. Starting to debug my physical serial interfaces I just couldn't find a way to get this to work. Setserial tells me the following: /dev/ttyS0, UART: 16550A, Port: 0x0000, IRQ: 4 /dev/ttyS1, UART: 16550A, Port: 0x0000, IRQ: 5 /dev/ttyS2, UART: 16550A, Port: 0x0000, IRQ: 7 /dev/ttyS3, UART: unknown, Port: 0x02e8, IRQ: 3 dmesg | grep tty this: [ 0.000000] Command line: BOOT_IMAGE=/vmlinuz-4.9.0-8-amd64 root=/dev/mapper/pen--01--vg-root ro console=ttyS0,115200n8 console=tty0,115200n8 [ 0.000000] Kernel command line: BOOT_IMAGE=/vmlinuz-4.9.0-8-amd64 root=/dev/mapper/pen--01--vg-root ro console=ttyS0,115200n8 console=tty0,115200n8 [ 0.000000] console [tty0] enabled [ 0.000000] console [ttyS0] enabled [ 49.308213] console [ttyS0] disabled [ 49.312635] dw-apb-uart.0: ttyS0 at MMIO 0x81424000 (irq = 4, base_baud = 115200) is a 16550A [ 49.407713] usb 1-8.2: Qualcomm USB modem converter now attached to ttyUSB0 [ 49.410489] usb 1-8.2: Qualcomm USB modem converter now attached to ttyUSB1 [ 49.411349] usb 1-8.2: Qualcomm USB modem converter now attached to ttyUSB2 [ 53.577143] console [ttyS0] enabled [ 54.744793] dw-apb-uart.1: ttyS1 at MMIO 0x81422000 (irq = 5, base_baud = 115200) is a 16550A [ 54.755771] dw-apb-uart.2: ttyS2 at MMIO 0x81420000 (irq = 7, base_baud = 115200) is a 16550A [ 115.902885] ttyS2 - failed to request DMA Why is ttyS0 disabled and again reenabled? How can I enable it/use it properly? How can I get my login shell on serial? What am I missing?

I have a headless server setup, and when I ssh in I get

: untrusted X11 forwarding setup failed: xauth key data not generated

Everything still works fine, since there's no X11 data to forward. But is there a way to modify the ssh config on the server to silence this warning?

I followed the steps from the OpenVPN Access Server Portal to install OpenVPN Access Server on an Ubuntu Server 22 headless VM. After installing and automatically configuring the service, I went to log in at the admin portal and once I did I was presented with the license agreement, only when I click Agree it just refreshes the page and does nothing. So I went back and manually reconfigured the service with the /usr/local/openvpn_as/bin/ovpn-init tool and I was presented with the exact same issue. I looked through the logs and I couldn't find anything obvious there either, but I'll post them here if anyone thinks it will help. Am I doing something wrong here or is my installation just broken?

I am half mad and I suddenly had an idea. I don't care that it is very impractical and will cause me trouble, I want it for some reason. I want to have Linux, without a desktop environment, but be able to run certain apps in a GUI, for example viewing images or video's, or playing games on steam. If there is some kind of a desktop environment, that just looks like a terminal when not doing anything that requires a gui, I will be fine with that.

I'm trying to set up a headless FreeBSD 13.1 system. I've done so successfully before on this particular machine, but for whatever reason, after a complete reinstall, it doesn't want to boot without a monitor attached via HDMI. Since it never shows up on the network, I'm assuming it never even gets past the boot prompt. It boots as expected if I plug in an HDMI cable to a monitor before powering up the machine. Plugging in a monitor after unsuccessfully trying to boot without a monitor does not work (black display). I tried adding -P to /boot.config, but this had no effect. The system is, apart from that, fairly vanilla. The system is a fan-less Chinese(?) "Basoaro" system with an American Megatrends BIOS, but I can't find anything in the BIOS that seems relevant, and since it *used* to be possible to boot the system without an attached monitor, I'm assuming it's not a BIOS-related issue. Whatever I did before to get it to boot headless, I've forgotten what it was.

Can you install VS-Code-CLI on a headless machine, without installing a desktop environment too? (I sometimes wish to access my code on an Android tablet, using the PWA version of Code that doesn't allow ssh, only the VS Code tunnel.) system spec:

No monitors, keyboards, speakers or anything.
Old Atom processor, 4G RAM
Debian 12 6.1.0-amd64 GCC 12.2
Node 20.1

Running code returns:

$ ./code 
No installation of Visual Studio Code stable was found.
Install it from your system's package manager or https://code.visualstudio.com , restart your shell, and try again.

If you already installed Visual Studio Code and we didn't detect it, run code version use stable --install-dir /path/to/installation

To install the Debian package code_1.85.1-1702462158_amd64.deb, depends on a display server, a window manager, a sound-subsystem, 1000s of libraries, etc. Which I don't need, or even want to attempt to install, on such a low-powered machine! The same would apply to Raspberry Pi headless systems too. The Debian package for code, depends on all these things. Has anyone tried something different, or could the Debian package be split into, for example, code-core into code-gui

I'm aware of Selenium python library or headless chromium webdriver, but those solutions are heavy and clunky to interface from a shell session (bash, dash, etc.). Is there a minimal and lightweight tool so I can do browsing automation like getElementById or click on buttons or run JavaScript functions utilizing the tool's cli from terminal? Note that we could curl then awk, jq, grep but again, these tools are not specialized for hypertext markup and using regexp for pattern matching HTML parameters is not advised generally. Additionally a simple HTML parser can't extract into from a Single Page Application webapp.