services setup often "fails" silently, making the debugging process very tedious. This can happen when there are similar containers on the system (e.g. same port/volume/names used for something else).

This kind of issue happen frequently (try to search for the postgres password authentication failed for user error string on the Discord server, most of the time it's related to a wrong infrastructure brought up by compose).

@ntarocco suggested "clean up" scripts are very useful in this situation and I usually reach a working situation with those, but that also mean that any other (non invenio related) containers running locally get destroyed.

This may be done in / fixed by #355