Since the Syncthing DB structure was modified, my Syncthing docker container (Linuxserver.io ) malforms the database with no apparent reason (server shutdown, etc.) - again and again. Each time, I have to (recommended fix) remove the index folder, re-scan, and remove all the sync-conflicts that have piled up until I notice the problem again.

For a productive server, this issue is really more than annoying... According to the Syncthing forum, this might be related to the docker containers (both Linuxserver.io and binhex) using the SQLite. The SQLite Go backend seems not really well tested, so the recommended solution is to use the original Syncthing package (or something closer to it) instead.

Is there a docker container for Syncthing with a non-SQL backend? Or is it possible to install the original Syncthing package on Unraid?

I'm not aware of syncthing being able to use anything other than sqlite?

Anyway I've been using the binhex container for 4 years and never had a single corruption event.

There must be something wrong with your system.

https://docs.syncthing.net/users/faq.html#my-syncthing-database-is-corrupt

I suggest running a memtest from the boot screen for 24 hours to rule out the potential for bad ram.