Open kuhball opened 3 months ago
Happy to merge PRs fixing the problem!
I ran into the same issue too but I'm not using podman. The error I get is:
failed to register layer: failed to Lchown "/pretalx/src/pretalx/frontend/schedule-editor/node_modules/character-parser/.npmignore" for UID 1618591017, GID 1876110778 (try increasing the number of subordinate IDs in /etc/subuid and /etc/subgid): lchown /pretalx/src/pretalx/frontend/schedule-editor/node_modules/character-parser/.npmignore: invalid argument
In my case I am running docker under a Linux container (LXD) and I think it has similar subuid limitations as podman.
I don't think it's anything with the code, but something about how this is being built because I can use the scripts as is to build the docker image.
Push it to my org nexus docker registry and then pull it down from another machine fine and use docker-compose.yml as usual.
After a bit of searching this seems to be a npm problem - https://github.com/npm/cli/issues/5889 . The used python:3.10-bookworm
ships with npm version 9.2.0 which has been released before the issue has been closed.
Solution to this would probably be a newer version of npm, not totally sure what exact npm version fixes this.
pretalx only supports currently supported Nodejs versions, and 9.2.0 is way outdated at this point – the oldest still-maintained version is v18 at the moment.
I was talking about the npm version, not nodejs. Current npm version is 10.5.1. The mentioned npm version is the one provided by the bookworm repositorys.
Ah, gotcha, sorry!
I'm trying to get pretalx to run within rootless podman. Pulling the image from dockerhub the following issue occurs:
In rootless podman the user normally gets a start uid and a range - f.e.
core:100000:1000000
. Looking at the requested id's this seems unnecessarily high and far from each other. Is there a particular reason for this?Workaround is to build the image locally within rootless podman.