Description of problem:

I recommend to make the RHCOS images cache as the systemctl service since when we reboot the bastion server, the container does not come up automatically and it may cause outage at live systems under some circumstances.

Ex:

podman generate systemd \
--new --name rhcos_image_cache \
> /etc/systemd/system/rhcos_image_cache

Version-Release number of selected component (if applicable):
4.7

How reproducible:

Steps to Reproduce:
stop rhcos_image_cache container
upgrade the ocp cluster
metal3 pods does not come up as they can not retrieve the images from podman container

Actual results:

Expected results:

Additional info: