Running the windows installer fails
Also isnt recognized by windows screener which might be part of the problem in some way.
Running the windows portable fails
Unsigned bat may be part of the problem.
This also flags as APT Behavior T1055 Defense Evasion
It was rerun with the entire folder as well as executables exempted and it still failed.
This site can’t be reached
127.0.0.1 refused to connect.
Try:
-
Checking the connection -
Checking the proxy and the firewall
ERR_CONNECTION_REFUSED
Running it via docker with ubuntu fails
The exposed ports dont load despite numerous variations attempted.
Running it via a docker that is supposed to be specifically for comfyui (https://hub.docker.com/r/mmartial/comfyui-nvidia-docker) fails in numerous different ways including:
========================================================= Starting script (ID: 1)== Running comfyui-nvidia_init.bash in / as comfytoo
-
script_fullname: /comfyui-nvidia_init.bash– WANTED_UID: “1000”– WANTED_GID: “1000”– SECURITY_LEVEL: “normal”– BASE_DIRECTORY: “/basedir”== Most Environment variables set== Extracting base image information– Base image details (from /etc/image_base.txt):DOCKER_FROM: nvidia/cuda:13.0.0-cudnn-devel-ubuntu24.04CUDNN: libcudnn9-cuda-13 (9.12.0.46-1)COMFYUI_NVIDIA_DOCKER_VERSION: unknown– COMFYUIUSER_DIR: “/comfy”– BUILD_BASE: “unknown”!! ERROR: Invalid BUILD_BASE value!! Exiting script (ID: 1)
========================================================= Starting script (ID: 1)== Running comfyui-nvidia_init.bash in / as comfytoo
-
script_fullname: /comfyui-nvidia_init.bash– WANTED_UID: “1000”– WANTED_GID: “1000”– SECURITY_LEVEL: “normal”– BASE_DIRECTORY: “/basedir”== Most Environment variables set== Extracting base image information– Base image details (from /etc/image_base.txt):DOCKER_FROM: nvidia/cuda:13.0.0-cudnn-devel-ubuntu24.04CUDNN: libcudnn9-cuda-13 (9.12.0.46-1)COMFYUI_NVIDIA_DOCKER_VERSION: unknown– COMFYUIUSER_DIR: “/comfy”– BUILD_BASE: “unknown”!! ERROR: Invalid BUILD_BASE value!! Exiting script (ID: 1)
– Found Max driver CUDA version: 13.0 (major: 13, minor: 0)– Found Max container CUDA version: (major: , minor: )/comfyui-nvidia_init.bash: line 295: [: 0: unary operator expected
== Testing write access as the comfy user to the run directory!! ERROR: Directory /comfy/mnt owned by unexpected user/group, expected 1000:1000, actual 0:0 – FORCE_CHOWN will not work for this folder, it is a PATH mounted at container startup and requires a manual fix: chown -R 1000:1000 foldername!! Exiting script (ID: 49)!! ERROR: subscript failed!! Exiting script (ID: 1)
I havent been able to find any fixes anywhere. No LLM that I have tried has been able to fix anything.
I have used many LLM and other products without anywhere close to this many issues and then still not being able to actually run this.
