Hey all,
I am trying to create a workflow that clones a repository when a webhook gets executed.
Except, when the Git Clone node gets executed, it just gets stuck without any feedback (error, log, or result), the specified repository folder does not get populated either.
A similar thing happens when using the “Read File(s) From Disk” node. When I configure this node to read a specified file, it also gets stuck.
What is the error message (if any)?
Please share your workflow
(Select the nodes on your canvas and use the keyboard shortcuts CMD+C/CTRL+C and CMD+V/CTRL+V to copy and paste the workflow.)
Hey @ewoud Welcome to the n8n Community
A clone that hangs with no error and no folder is almost always git waiting on a prompt it can’t show you. Inside the container there’s no terminal, so a private repo asking for credentials, or an SSH URL asking you to accept an unknown host key, blocks forever instead of failing. Set GIT_TERMINAL_PROMPT=0 on the container and run it again. You’ll get a real error in seconds, which tells you which of the two it is.
The file node stalling is worth treating as its own problem. Check the folder is actually bind mounted into the container and not just a UGOS share path, since the node sees the container filesystem and nothing else.
Hi @ewoud Welcome!
Both nodes touch the path before they do any real work. The Git node resolves the repository path with realpath before it shells out to git, and the file node runs the glob before it opens anything, so neither one reaches the operation you configured. A path that is missing, unreachable from the container, or outside the allowed directory fails immediately with an error rather than hanging, so a spinner that never resolves and never throws means the filesystem call itself is not returning.
That happens when the path lands on a mount that is still attached but no longer answering, which is what a remote or cloud folder on the NAS does once its backing connection drops. Check the same path from inside the container:
docker exec -it <your-n8n-container> ls -la /your/configured/path
If that sits there too, the problem is under the mount and n8n can only block on it. Point both nodes at a plain Docker volume instead. On 2.x the path also has to sit inside N8N_RESTRICT_FILE_ACCESS_TO, which now defaults to ~/.n8n-files.
The points above highlight the exact two reasons why Git Clone or File nodes hang indefinitely without throwing an error in n8n Docker instances:
Git waiting for interactive credentials:
When cloning via SSH or HTTPS without pre-accepted host keys, Git freezes waiting for user input. Setting this environment variable forces Git to fail fast and output the true error message:
GIT_TERMINAL_PROMPT=0
Stale or hung network mount:
If your target directory resides on a network share (NAS, SMB, NFS) that has disconnected, system I/O freezes. Verify whether your container can read the path by running:
docker exec -it <container_name> ls -la /path/to/target
File access restrictions in n8n v2.x+:
Make sure the destination folder is whitelisted in your environment variables:
N8N_RESTRICT_FILE_ACCESS_TO=/path/to/target
Testing these three items will pinpoint whether it is a credential prompt lockup or a Docker mount issue.
