After restarting the server, it almost forgot all the containers!

Normal updating replaced something that required a reboot. It took a little bit, but came back after just a few minutes longer than I would have liked. I think it had to check a disk, but it booted fine. I did the usual and rebuilt the firewall. Then I tried to check on the containers. No containers!

Previously, and on all the other servers, Docker (or Podman, but Docker in this instance) remembers all of the containers and restarts those so configured, but none were there!

I ran through the shell history and restarted all of those that I had started that way. Then I peeked at the Portainer.io container that I'd just reconfigured prior to doing the restart. It saw all the containers started from the command line (which did include Portainer). And it also had all of the Portainer Stacks, which is essentially its management of Docker Compose scripts. I clicked into each and restarted them, also.

During the reboot, all of my online monitoring triggered that the sites were online. As I ran through the shell history and started the stacks, they started pinging that they returned to the Internet. I ran through a quick browser and curl check, watching the reverse-proxy logs to make sure that the pages were loading from the server, and it seems no worse for wear.

I'm rethinking my infrastructure, and reconsidering Ansible. I used to love Ansible at work, until they did better things, but never integrated it into my lab. That it runs as if from a shell command, and the entire infrastructure is contained in a Git[Lab] repository, makes it pretty easy to reproduce installs and updates, even "from scratch."

But, for now, the things are working.