Docker Desktop GUI is not starting under Debian 10 (Plasma or Gnome)

[GrimzEcho]

• [x] I have tried with the latest version of Docker Desktop
• [x] I have tried disabling enabled experimental features
• [ ] I have uploaded Diagnostics
• Diagnostics ID: I am not able to get to a point where I can run diagnostics, and the linked documentation in the issue template is for Windows.

Expected behavior

When I launch Docker Desktop from the desktop application menu (or when starting it using systemctl --user start docker-desktop), the GUI application starts and opens

Actual behavior

The GUI application flashes a brief icon on the taskbar, then closes. No application window opens, and no errors are reported (either via the GUI or the log file)

Information

• Linux distro: Debian
• Distro version: Debian 10 (buster)
• Desktop:
  • KDE Plasma (5.14.5)
    • Qt: 5.11.3
  • Gnome (3.30.2)
  • Gnome Terminal: 3.30.2
• Docker Desktop Version: 4.11.0-amd64

Steps to reproduce the behavior

1. Follow the instructions from https://docs.docker.com/desktop/install/debian/
2. Ensure that the correct repository was added, and that the install script completed without errors
3. Confirm that Gnome Terminal is installed
4. Confirm that the install script completed without errors
5. Attempt to run Docker Desktop from Plasma using its application entry (Plasma places it in the Lost & Found category)
6. Note that the whale icon briefly appears on the taskbar as if an application was starting, but disappears within a second
7. Run the command systemctl --user start docker-desktop
8. Note that no GUI application opens
9. Restart and log into Gnome desktop
10. Install the AppIndicator and KStatusNotifierItem Gnome Extension
11. Attempt to start Docker Desktop using the application entry
12. Note the same behavior as step 6
13. Remove ~/.docker directory
14. Restart and try again in KDE Plasma
15. Same behavior as step 6

Attempted fixes

• Uninstall all docker related packages (via apt list --installed | grep docker)
• Remove ~/.docker directory
• Install docker engine before installing docker desktop
• Install docker engine after installing docker desktop (this is the current state of my system, as I still need to be able to host docker containers)

Relevant logs

• .docker/log/host/docker-desktop.log

-------------------------------------------------------------------------------->8
-------------------------------------------------------------------------------->8
[2022-09-08T18:51:59.247817247Z][docker-desktop][I] running under systemd
-------------------------------------------------------------------------------->8
-------------------------------------------------------------------------------->8
[2022-09-08T20:00:07.941991881Z][docker-desktop][I] running under systemd

 

• last few lines of .docker/log/host/com.docker.bankend.log

[2022-09-08T20:01:48.106006384Z][com.docker.backend][W] 5469847e-PauseHDL /pause/events server not replying: Get "http://ipc/pause/events": dial unix procd.sock: connect: no such file or directory
[2022-09-08T20:01:48.106006624Z][com.docker.backend][W] still waiting to update HTTP proxy on http-proxy-control.sock after 1h16m2.42737794s. Last error was Post "http://localhost/vm/httpproxy": dial unix http-proxy-control.sock: connect: no such file or directory
[2022-09-08T20:01:58.111412229Z][com.docker.backend][W] still waiting to update HTTP proxy on http-proxy-control.sock after 1h16m12.432770044s. Last error was Post "http://localhost/vm/httpproxy": dial unix http-proxy-control.sock: connect: no such file or directory
[2022-09-08T20:02:08.114823453Z][com.docker.backend][W] still waiting to update HTTP proxy on http-proxy-control.sock after 1h16m22.436194348s. Last error was Post "http://localhost/vm/httpproxy": dial unix http-proxy-control.sock: connect: no such file or directory
[2022-09-08T20:02:18.119770151Z][com.docker.backend][W] 5469847e-PauseHDL /pause/events server not replying: Get "http://ipc/pause/events": dial unix procd.sock: connect: no such file or directory
[2022-09-08T20:02:18.119779541Z][com.docker.backend][W] still waiting to update HTTP proxy on http-proxy-control.sock after 1h16m32.441147296s. Last error was Post "http://localhost/vm/httpproxy": dial unix http-proxy-control.sock: connect: no such file or directory

 

• output of systemctl --user status docker-desktop

docker-desktop.service - Docker Desktop
   Loaded: loaded (/usr/lib/systemd/user/docker-desktop.service; disabled; vendor preset: enabled)
   Active: active (running) since Thu 2022-09-08 12:45:45 MDT; 1h 17min ago
 Main PID: 5252 (com.docker.back)
   CGroup: /user.slice/user-1001.slice/[email protected]/docker-desktop.service
           ├─5252 /opt/docker-desktop/bin/com.docker.backend --with-frontend
           └─5272 /opt/docker-desktop/bin/com.docker.backend --with-frontend

 

• .docker/logs/vm -- directory is empty

 

• apt list --installed | grep docker (this is after installing docker engine and docker-compose-plugin). The issue occurs even when docker engine and docker-compose-plugin are not installed

docker-ce-cli/buster,now 5:20.10.17~3-0~debian-buster amd64 [installed]
docker-ce-rootless-extras/buster,now 5:20.10.17~3-0~debian-buster amd64 [installed,automatic]
docker-ce/buster,now 5:20.10.17~3-0~debian-buster amd64 [installed]
docker-compose-plugin/buster,now 2.6.0~debian-buster amd64 [installed]
docker-desktop/now 4.11.0-83626 amd64 [installed,local]
docker-scan-plugin/buster,now 0.17.0~debian-buster amd64 [installed,automatic]

[aiordache]

Hi @BrianHVB Can you check the file ~/.docker/desktop/log/host/com.docker.driver.amd64-linux.stderr.log for errors logged on VM startup please?

Try to run the diagnose tool from the terminal as documented in https://docs.docker.com/desktop/troubleshoot/overview/#diagnose-from-the-terminal to create and upload a diagnostics bundle.

[GrimzEcho]

@aiordache

There are no stderr logs in ~/.docker/desktop/log/host, just stdout. None of the three log files shown below contain any recognizable errors.

My dianostics id is: b40cb4d9-bed0-42eb-971b-0f7d5c2eb788/20220912145043

[aiordache]

@BrianHVB From the self diagnose traces, it looks like you haven't enabled kvm:

failures detected
1 : does the host support virtualization?
    Failed with: stat /dev/kvm: no such file or directory
The Docker engine runs inside a Linux VM. Therefore the host must support virtualization.

Check that hardware-assisted virtualization (either Intel VMX or AMD SVM) and Data Execution Prevention (sometimes labeled XD or Execute Disable or NX) are enabled in your BIOS.

Check your bootloader is configured to launch Hyper-V.

See https://docs.docker.com/desktop/windows/troubleshoot/#virtualization

Please follow instructions at https://docs.docker.com/desktop/install/linux-install/#kvm-virtualization-support to setup kvm.

We will update the link and description in the self-diagnose for the linux platform.

[GrimzEcho]

@aiordache

Thank you. I have enabled virtualization support in my bios and confirmed that it is loaded (I previously misunderstood this step). However, I am still experiencing the same symptoms (the Docker Desktop taskbar icon briefly flashes, but the application never loads)

The updated diagnostic ID is 70fff107-822f-47cb-9c9f-5663f267a8aa/20220914152539.

Based on the self-check utility, two things failed:

1. vm has not started: failed to open kmsg.log: open /home/bhoule/.docker/desktop/log/vm/kmsg.log: no such file or directory
2. CLI context is set to docker-ce engine

I am unsure of the cause of the VM not starting. As for the CLI context, docker context ls shows only a single context at CLI context is set to docker-ce engine. There is no context for docker-desktop.

[aiordache]

@BrianHVB Seems like the backend may be stuck waiting for you to accept the license agreement. The license page should have been displayed at this point.

You can try to accept the license by directly setting the licenseTermsVersion in the settings file as below:

$ cat ~/.docker/desktop/settings.json | grep license
  "licenseTermsVersion": 2,

Then try to start it again but first make sure the com.docker.backend process is not running anymore.

[GrimzEcho]

No joy -- same symptoms after hard-setting licenseTermsVersion to 2 in settings.json.

Although I do appear to now at least have a new docker context set for docker desktop.

At this point, I think I should try and see if this is reproducible on a stock Debian 10 + KDE Plasma install. I know GH Issues is not the best place for support-style threads. It seems pretty clear that the VM is crashing w/o any direct loggable cause, but whether that is related to hardware, my specific software environment, or something more general I don't know.

[docker-robott]

There hasn't been any activity on this issue for a long time. If the problem is still relevant, mark the issue as fresh with a /remove-lifecycle stale comment. If not, this issue will be closed in 30 days.

Prevent issues from auto-closing with a /lifecycle frozen comment.

/lifecycle stale

[stueken]

I have a similar problem, but the self check tells me that docker is not running:

Please investigate the following 1 issue:

1 : The test: is the application running?
    Failed with: Docker is not running

Is the Docker Desktop application running?

However, if I run systemctl --user status docker-desktop.service, docker seems running, just the GUI won't show up:

● docker-desktop.service - Docker Desktop                                                                                                   
     Loaded: loaded (/usr/lib/systemd/user/docker-desktop.service; disabled; preset: enabled)                                               
     Active: active (running) since Fri 2023-03-17 15:48:21 CET; 9min ago                                                                   
   Main PID: 2348492 (com.docker.back)                                                                                                      
      Tasks: 70 (limit: 18669)                                                                                                              
     Memory: 5.4G                                                                                                                           
        CPU: 1min 14.508s                                                                                                                   
     CGroup: /user.slice/user-1000.slice/[email protected]/app.slice/docker-desktop.service                                                 
             ├─2348492 /opt/docker-desktop/bin/com.docker.backend                                                                           
             ├─2348618 com.docker.driver.amd64-linux -addr fd:3 -debug                                                                      
             ├─2348619 com.docker.extensions -address extension-manager.sock -watchdog                                                      
             ├─2348620 com.docker.dev-envs                                                                                                  
             ├─2348669 /opt/docker-desktop/bin/virtiofsd --socket-path=virtiofs.sock0 -o cache=auto -o source=/home --sandbox=none --announ>
             └─2348675 qemu-system-x86_64 -accel kvm -cpu host -machine q35 -m 3919 -smp 4 -kernel /opt/docker-desktop/linuxkit/kernel -app>
                                                                                                                                            
Mär 17 15:57:21 x1g7 com.docker.backend[2348492]: [2023-03-17T14:57:21.747413172Z][com.docker.backend][I] (de1de5e3) 5020642d-VMDiskUsageCl>

My Diagnostics ID: 799335d0-0618-4408-af0f-8dae8153aed2/20230317145434 (uploaded)

[HotScottPtII]

I have the same issue using Parrot OS (Debian based) Plasma desktop. Using Docker-desktop version 4.20.0 although I found the same problem using an earlier version

[j-shpartko-tqgg]

Here is the solution which is work for me (the latest Debian): https://github.com/docker/desktop-linux/issues/27