Really appreciate all the work here - just added a small monthly contribution to help out.
I couldn’t find a discussion on this elsewhere (pardon me if I missed it), but it would be cool to utilize Docker’s healthchecks: https://docs.docker.com/engine/reference/builder/#healthcheck
Then there could be a LSIO container that would auto heal zombie containers, maybe similar to something like this: https://github.com/willfarrell/docker-autoheal
There are some downsides to automatically restarting containers, but if the auto heal container is set up to only auto heal certain containers, I think you can overcome any issues. The tricky part is that there would have to be health checks added to every image. I did notice that a lot of images from publishers included a health check for their application. It would just need to be standardized across the fleet.