Running unmodified docker-compose example yields errors

Questions
1

Hello Friends:

Are you able to run the self-hosted docker-compose-local.yaml example provided here without error (whether you specify the debian or alpine image - it really doesn’t matter).

Why? After experiencing a variety of EACCESS-related errors with the various supplied docker images (including one I posted yesterday here), I decided to simply try the above without modification.

Sadly, I still get EACCESS-related errors coming from inside the container. And, again, I intentionally didn’t modify anything (except trying debian and then alpine image variants).

This time, one errors was:

user@fedora$ podman-compose -f ./docker-compose-unmodified.yaml up
54602a1ba94b64d599555999a63c362502898e9d555bd51c3a9be4b2091dd704
673f3f1ffcc7e7e2ba5e4edcd704c852a764c06840735137ec66febb2696d29b
[n8n-python] | UserSettings were generated and saved to: /home/node/.n8n/config
[n8n-python] | Error: EACCES: permission denied, open '/home/node/.n8n/config'

Are you able to run the provided example above, without modification, successfully?

To me it seems like something is not quite correct with the provided python enhancing images (i.e., debian and alpine).

Thank you! :blush:

2

It looks like your topic is missing some important information. Could you provide the following if applicable.

  • n8n version:
  • Database (default: SQLite):
  • n8n EXECUTIONS_PROCESS setting (default: own, main):
  • Running n8n via (Docker, npm, n8n cloud, desktop app):
  • Operating system:
3

Hey @nmvega , why do you want to use that (unofficial) image? I believe it was created prior to native support of Python in Code nodes. Python scripts are now supported natively, Using the Code node | n8n Docs.

2 Likes
4

Hi @ihortom

That’s a great question. Because I’m a Noob and just figuring things out. LoL. :smile: But that might explain the Docker image errors (i.e., code rot).

But seriously, I read that Python libraries are confined to what is available here (pyodide).

That library coverage does seems pretty comprehensive, though. There are many. I was just trying to get ahead in case one wasn’t available (say, PyTorch - though I probably won’t need that). But just anticipating.

Another reason is the problem I’m having here with the official image, which I posted here (yesterday). I had hoped another image might skirt that issue, but it only led to this new one.

I’ll just continue working with the officially released one, as you suggest.

Thank you for your reply!

5

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.