Production deploy fallback documentation

[aequitas]

In case of Github downtime, we need documentation on how to perform a manual build, test and deployment on a production server in case hotfixes need to be applied.

[bwbroersma]

The unregistry project describes it well:

The problem

You've built a Docker image locally. Now you need it on your server. Your options suck:

• Docker Hub / GitHub Container Registry - Your code is now public, or you're paying for private repos
• Self-hosted registry - Another service to maintain, secure, and pay for storage
• Save/Load - docker save | ssh <remote server> docker load transfers the entire image, even if 90% already exists
  on the server
• Rebuild remotely - Wastes time and server resources. Plus now you're debugging why the build fails in production

You just want to move an image from A to B. Why is this so hard?

The project is a solution to ship single layers (instead of save/load), maybe using their container and the 'pussh' alias is even simpler than remote building? But maybe there are more downsides to it I don't see now.