That doesn’t represent much. It just copies the server files, without the data folder but with all the apps and the config directory in the data/updater-INSTANCEID/backups/ directory in a folder with the name schema
So for example:
Unixtime is the seconds since January 1, 1970.
Converting Unixtime into human-readable time goes like this:
date -d "@1702631751" "+%Y-%m-%d %H:%M:%S"
so 1702631751 corresponds to 2023-12-15 10:15:51
However, no database dump is created without which you cannot undo anything.
Without a database backup and (not that important) a backup of the appdata_INSTANCEID folder, you cannot revert your update to the state from before the update.
For me it is a nice to have. I can fast compare versions on my machine with those backups but I create a backup from serverfiles, database and appdata together with some other important configuration files from acros the server before each update myself. The backup and restore steps are all perfectly scripted and stored in my .bash_aliases file. So my update steps looks like this:
When I then have problems with the new version I do
which shows me a list from all available snapshots I made where I select the snapshot I want to go back to, start restore job and after a minute it is all the old state again.
It is extremely important to have these steps perfectly organized and tested well, so that you know exactly what the steps are to be done, because when something goes wrong and you start to stumble around and encounters dificulties while your server is offline, maybe with an internal server error, that’s not a nice experience