So log files is signalling nginx to reLOAD. granted. But nginx is not reLOADING on this setup it is killing it. forcing the OP to perform a restart. (in Virtualmin or with systemctl) or by delaying the reLOAD.
As I said what is making this happen on this system and not on every nginx installation?
This appears to be related to a missing parameter on the new Debian 12 support. The reload isn’t instantaneous, hence the
sleep 5 parameter. I’m assuming from what is listed above is that reloading while a reload is in process is causing a problem and simply killing nginx. Notice
sleep 1 didn’t work.
In Log rotate the “Commands to run after rotation” was set to “systemctl restart nginx” so I added on that line “; sleep 5” (I did this after reading your post in this for Apache).
(sorry for late reply, was out all day.)