2 Servers 2 Locations, both running HA, both quit working after recent server update

Configuration
21

Ran across this thread and stranger yet, running official Docker-ce I noticed that current version is 19.03 again now with a subversion of 3-0 for the officially supported Debian config.

My HA has been horribly unstable with both Supervisor and all Docker stuff crashing outright for a couple of days.

Restored to VM backup just to get it running this morning. Console wouldn’t even run “ha” saying the container wasn’t responding.

Attempting to look at logs I would sometimes see supervisor complaining that one of the core Docker IP addresses had stopped responding three times and then nothing and the box was completely dead.

22

Ahh I take part of the above back. It wants to upgrade Docker again. Not doing it. Ha!

23

Interesting … a couple of other docker containers, e.g. NextCloudPi, also have problems with docker versions 5.20.

24

Okay, I ran into a problem with the TasmoAdmin add-on. It won’t start, but it also won’t report to me the name of the image with the image error. It also won’t uninstall.

How can I manually override this and wipe it out so I can re-install it?

25

Resolved. I took a shot in the dark.

docker image pull hassioaddons/sonweb-amd

…which downloaded the latest, then attempted to start and that’s when I got the error letting me know that I needed the particular version:

hassioaddons/sonweb-amd64:0.13.1

Pulled that image and it started.

26

I Also lost my HA, rebuild everything from scratch and find that it happen again!! Containers were being removed by docker-ce!! READ THIS!!!

27

Docker broke mine as well. I did the following on Ubuntu to restore service:
apt-get install docker-ce=5:19.03.14~3-0~ubuntu-focal

28

New Supervisor (2020.12.7) fixes the Docker issue. You can upgrade to Docker 20 now.

In fact, when I updated Supervisor this morning, I got the “unhealthy installation” warning… I upgraded my Docker and the warning went away.