Best Practices for Monolith Repo and Container Image use with Molecule

[DanSibbernsen]

Hi! Been using Molecule for a few weeks now, and I'm slowly getting the hang of it. My use case is a monolith repo with many roles that I'm deploying to SUSE. I have a container image that now works for my purposes, but I also have cluttered my roles with repeated Molecule files (molecule.yml, Dockerfile.j2, prepare.yml, cleanup.yml) that are 100% copy / paste across all my roles

path/to/role/molecule/
└── default
    ├── cleanup.yml
    ├── converge.yml
    ├── Dockerfile.j2
    ├── molecule.yml
    ├── prepare.yml
    └── verify.yml

I started digging into the .config/molecule/config.yml and realized I could transplant almost all of my repeated yml files into a .config directory and use molecule --base-config, so then I ended up with these directory structures

.config/
└── molecule
    ├── cleanup.yml
    ├── config.yml
    ├── Dockerfile.j2
    └── prepare.yml

/path/to/role/molecule/
└── default
    ├── converge.yml
    ├── molecule.yml -- empty file, apparently molecule still requires it
    └── verify.yml

Everything except the Dockerfile.j2 seemed to work fine. Digging into PR51 and the 1.0.1 Discussion, I figured the changes that would make transplanting the dockerfile into another directory were probably counter-intuitive to this module, hence why those changes were removed.

So now I'm wondering: for a monolith repo that uses the same container image over and over again, is it best to just create a container image beforehand and have molecule use that? As opposed to having molecule create the image?

Any feedback on my approach is much appreciated.