Unfortunately, this also happened to me a few weeks ago.
Accidentally discovered that the container of my backup witness node was no longer running. When I tried to restart, I got an error message that the database was corrupt. I then had to re-download the data as well.

As a result, I started looking around for monitoring apps. I have now set up check_mk (https://checkmk.com/) and monitor all servers and even every container...

I haven't tried all the options yet, but I now get notifications and can perform various actions. The next thing will be the automatic switching of Witness nodes with the help of the app. This is currently done by a script that only checks the availability of the servers, but not the containers themselves...