Docs should tell how to properly install external modules

[hadmut]

Hi,

would be nice if the documentation could tell something about what the intended way of installing external modules is.

I'm currently building docker images based on container-openvoxserver with

ARG PFpath=/etc/puppetlabs/code

ADD Puppetfile ${PFpath}

RUN mkdir -p /opt/puppet/modules &&
cd ${PFpath} &&
r10k puppetfile check &&
r10k puppetfile install

but that obviously wouldn't work with podman's autofresh.

It would also be possible to add a script to container-entrypoint.d/ to load them at boot time, but that's also somewhat homebrewn and might be incompatible with openvox's future strategy.

So what is the clean, reliable, and autorefresh-stable way to install external Modules?

I would have expected that a docker image itself looks at start time wether there is a Puppetfile and runs r10k. So I'm in doubt what openvox's future stategy will be, and how to load modules without breaking future strategies.

regards

[rwaffen]

good point! this could really be improved. the way I think this should be done is like this: https://github.com/voxpupuli/crafty/tree/main/openvox/r10k

[rwaffen]

the plan is, to have a external r10k container and a shared volume. then run the r10k maybe time triggered or via webhook. the openvoxserver container then will get the code via the shared volume. the future plans for this container will be to remove the r10k binary. maybe. i'm not 100% sure about this, yet.