Are you getting failures to graceful start of Apache at logrotation?
Something like this in your logs?
[core:warn] [pid 1234:tid 123456789012345] AH00098: pid file /var/run/apache2/apache2.pid overwritten -- Unclean shutdown of previous Apache run? [mpm_event:notice] [pid 1234:tid 123456789012345] AH00489: Apache/2.x.xx (Ubuntu) mod_fcgid/x.x.x OpenSSL/x.xx.xxx configured -- resuming normal operations ...[mpm_event:notice] [pid 1234:tid 123456789012345] AH00493: SIGUSR1 received. Doing graceful restart [core:notice] [pid 1234] AH00060: seg fault or similar nasty error detected in the parent process [fcgid:error] [pid 1235:tid 123456789012346] FastCGI process 1235 still did not exit, terminating forcefully
Looks like some changes were made that don't get re-applied to older Virtualmin installs (2018 era and before):
https://www.virtualmin.com/node/22024
I think I might have been here before and fixed it with this:
blog/01328216702
See the "Q: Why are the apache logs not rotating?" bit....
Looks like the combined logrotate with many log lines in it is the "fix" for now.
If you already had that and are still having issues, try to bump up the sleep time out.
Ok, update time. Got sick of babysitting this thing and added a cron job with this:
apachealive.py
Bit of a hack, but seems to be "working" for now, and allow for minimal downtime in the wee hours of the night!