Docker prune old images

[aaron97neu]

Every time container-maintainer pulls down a new image, the old one (and all unused volumes, networks, etc) remain and take up space. As development progresses, this will cause space issues in our oldest production devices. A docker system prune should be called regularly to clean this up.

The challenge comes in that we would like to be able to rollback changes quickly if a new docker container does not work properly, so simply calling docker system prune at the end of the container-maintainer script could introduce new issues when a container immediately fails (We would have to redownload the old container). As @abalmos mentioned, only pruning images older than a week should be sufficient

This would likely live in container-maintainer

[aaron97neu]

Looks like there is an option to filter what prune
Full systemd integration:
https://gist.github.com/pohmelie/272b5c2487df60eb7ef1c9db4936b331

More details
https://stackoverflow.com/questions/32723111/how-to-remove-old-and-unused-docker-images
https://docs.docker.com/engine/reference/commandline/system_prune/#filtering

Prob best to throw this at the end of container-maintainer instead of a separate systemd process